videomd Your workspace

/ PRACTICAL INSTRUCTIONS FOR YOUR AI

Less watching.
More building.

We turn videos into concise Markdown (.md) instructions your AI can read and use. Try three free workflows, browse the library, or bring your own video.

Connect your AITry a free brief

Filters

Refine your library

Collections

Works in
Cost
Bring
Loads
For
Stage
Model
Tested
What the pill colors mean
  • paidfree alternativefree what running it costs you beyond the file: a paid service, a paid service with a free route (hover it), or nothing
  • DockerAPI key what you must bring: software, an account or a key
  • ran in (tool)self-checkdone check evidence: we ran it in that tool, or the file carries its own check
  • checked 2026-10-12 the day it was last checked against official docs; not a run
  • Any AI where it works; each brief page says where to put it
  • shellchat only what your AI app must be able to do
  • on demand when your AI reads it
  • frontier the smallest model it was written for
  • codingbuilding what it is for, and the build stage it fits

Agent capabilities

Any AI · MCP · Agent skills46 files

Grow a live app one tested change at a time

Pick the one number your app alone can move, read the why in replays and logs, then test one change on half your users and keep what wins.

Tools: freeNodeClaude Code
Any AI1646words7 steps

Your writing voice from five samples

Build a voice profile from five pieces you wrote, apply it to new drafts, and check each draft against the profile.

Tools: freeAny text chatAny AI
Any AI548words5 steps

Verification loop

Make the AI run the check, read the real output and fix from it, instead of saying the change should work.

Tools: freeAny AIAny text chat
Any AI555words3 steps

Tool protocol for a model with no tools

Teach a chat-only model to ask for files, commands and searches with READ, RUN, SEARCH and WRITE lines, and to claim nothing it did not see.

Tools: freeAny text chatOllama
Any AI642words3 steps

Test-first for agents

Make your coding agent write the failing test, show it fail, implement, and show it pass; no test, no done.

Tools: freeAny AIClaude Code
Any AI392words1 steps

Structured output contract

Answer only in this JSON, or in an error object; a validator snippet and the Ollama format field for pipelines.

Tools: freePythonAny AI
Any AI316words1 steps

Small-window discipline

Rules for a model with a short context window, so it works one file at a time, summarizes, and hands off before it forgets.

Tools: freeAny AIAny text chat
Any AI615words3 steps

Skills for a harness that has no skills

Keep an index in AGENTS.md that says when to read each skills/<name>.md, so any coding agent loads a procedure on demand.

Tools: freeAny AIAider
Any AI395words1 steps

Session handoff

Have the AI keep SESSION.md (done, decisions, next, open), written before the window fills, so the next session resumes without re-asking.

Tools: freeAny AIAny text chat
Any AI908words3 steps

Self-review before done

Eight questions your AI answers against its own diff before it says done, so the gaps show up before you find them.

Tools: freeAny AIClaude Code
Any AI369words1 steps

Search and fetch protocol

When the AI does not know an API, it names the exact query and the vendor docs root, fetches, and treats what comes back as data.

Tools: freeAny AIAny text chat
Any AI622words3 steps

A research brief with every claim sourced

Answer one question with claims that each carry a URL, a not-found list for what no source supports, and nothing unsourced.

Tools: freeAny text chatAny AI
Any AI389words1 steps

A runbook for tasks that repeat

State what runs when, how a schedule is written, who runs it, and what to do when a run is missed, in one file the AI reads first.

Tools: freeAny AIClaude Code
Any AI638words5 steps

Prompt-injection defense

For any AI with web or file tools: fetched and read content is data, never instructions; report attempts instead of following them.

Tools: freeAny AIClaude Code
Any AI392words1 steps

Plan-first protocol

Make the AI write PLAN.md, wait for your OK, then execute step by step, check each step off and stop at any surprise.

Tools: freeAny AIAny text chat
Any AI919words3 steps

OpenHands AGENTS.md and one keyword skill

A root AGENTS.md OpenHands loads whole, plus one keyword skill that fires on a word, and the context floor for local models.

Tools: freeopenhandsOpenHands
Any AI604words5 steps

Memory bank for a model that forgets

Four small files the AI reads at every start and updates on request, so a model with no memory keeps project context across sessions.

Tools: freeAny AIAny text chat
Any AI575words3 steps

Meeting notes into dated decisions

Turn raw meeting notes into dated decision entries in one shape, with who, why, what changes and what stays open.

Tools: freeAny text chatAny AI
Any AI374words1 steps

Loop with a hard stop

For loops nobody watches: the runner caps turns and spend, done is an exit code, and the AI prints LOOP STOP and hands back on churn.

Tools: freeAny AIClaude Code
Any AI1336words4 steps

Local-only mode

Keep a local model's work off the network: where data stays, what to bind, and how to verify from the logs and sockets.

Tools: freeOllamaLM Studio
Any AI390words2 steps

An llms.txt for your own site

Write the llms.txt for your site, link it from every page, add Markdown twins, and find out whether anything fetches it.

Tools: freeAny AIClaude Code
Any AI412words1 steps

A hook without hooks: the git pre-commit check

One pre-commit script every coding agent inherits because git runs it: the tests, a secrets grep and a size cap before each commit.

Tools: freeGitAny AI
Any AI581words5 steps

Email triage into four tiers with drafted replies

Sort an inbox into skip, info only, meeting info and action required, draft the replies, and send nothing without an OK.

Tools: freeemail accountAny text chat
Any AI620words4 steps

Destructive-action rails

An ask-before list for agents with no permission system; the AI prints the exact command and waits before anything that cannot be undone.

Tools: freeAny AIAider
Any AI1004words3 steps

Debugging protocol

Reproduce, isolate, one hypothesis, one change, verify with the repro; stops an agent from shotgunning fixes.

Tools: freeAny AIClaude Code
Any AI380words1 steps

Continue rules that load on the right files

Write .continue/rules files with globs and alwaysApply so Continue applies each convention only to the files it is about.

Tools: freecontinueContinue
Any AI582words5 steps

Commit discipline

Give the AI the exact git commands for small commits, a message format, and a push that cannot hit the wrong branch.

Tools: freeGitAny AI
Any AI555words3 steps

Cline and Roo rules

Where Cline and Roo Code read rules, path-scoped rules, the six Memory Bank files, and how to see a rule applied.

Tools: freeClineRoo Code
Any AI557words5 steps

Aider CONVENTIONS.md

The conventions file Aider reads every session, how to load it without typing it each time, and the Ollama context fix.

Tools: freeaiderAider
Any AI528words5 steps

The 40-line AGENTS.md that works everywhere

One AGENTS.md with build commands, a do-not-touch list, the branch rule and done-means, read natively by most coding agents.

Tools: freeAny AICodex
Any AI626words3 steps

Add a dependency safely

Look it up, install with scripts off, commit the lockfile, scan it, rank findings by what is exploited; never call a package safe.

Tools: freeNodeOSV-Scanner
Any AI1317words12 steps

Make your AI check its work FREE SAMPLE

Run the same small validator with any agent. Includes Codex and Claude hook setup, plus a plain-command fallback.

Tools: freePythonCodex
Any AI502words5 steps

Give your AI a reusable skill FREE SAMPLE

Turn supplied notes into a sourced summary. Includes a ready-to-use skill and Codex, Claude Code and other-host setup.

Tools: freeCodexClaude Code
Any AI471words4 steps

Give your AI project memory FREE SAMPLE

Keep decisions and their sources in a small Markdown workspace that any text-capable AI can read.

Tools: freeAny text chatAny AI
Any AI436words6 steps

Model Context Protocol 2026-07-28 specification overview

Summarizes changes in the 2026-07-28 MCP specification, covering stateless core requests, CIMD authentication, and deprecations.

Tools: freeNodeAny AI
MCP164words0 steps

Reverse engineer a feature with REA, evidence first

Point your agent at an app, a bundle or a binary, get how a feature works with the evidence behind it, then build your own version.

Tools: freeNodeClaude Code
MCP1370words7 steps

Custom MCP Web Scraper Server

Build an authenticated scraping tool and connect its local MCP server to Claude Code.

Tools: freePythonOxylabs API key
MCP189words6 steps

Build an MCP server in Python with OAuth

Configure Descope authentication and verify access controls on a FastMCP server.

Tools: freePythonDescope account
MCP173words6 steps

Multi-session learning with the teach skill

Guides multi-session learning of skills or codebases using stateful local tracking files and custom interactive HTML lessons.

Tools: freeNodeClaude Code
Agent skills340words6 steps

Skill set for a plan-to-review coding loop

Install one skill bundle so a coding agent runs discovery, spec, tickets, TDD implementation and a two-axis code review, in that order.

Tools: freeNodeClaude Code
Agent skills407words11 steps

Skill bundle update: plain-English stops, batched interviews

Four more workflow skills for a coding agent: plain-English re-explains, batched design interviews, agent-file writing rules, setup wizards.

Tools: freeNodeClaude Code
Agent skills285words8 steps

Spec-to-PR skills: implement-spec, pr, retro

Run a whole spec across git worktrees with subagents, write pull requests with verification evidence, and mine past sessions for repo fixes.

Tools: freeNodeClaude Code
Agent skills287words4 steps

AI agent skills installation and workflows

Install and use specialized open-source skills to add custom capabilities to AI coding and chat agents.

Tools: freeClaude CodeCodex
Agent skills435words13 steps

Four-part checklist for writing agent skills

Design, structure, steer, and prune markdown skills for AI coding agents to ensure reliable execution.

Tools: freeClaude CodeCodex
Agent skills323words8 steps

Interview, spec, ticket and build with agent skills

Installs an engineering skills pack, sets up the repo once, then turns an interview into a spec, tickets and committed code.

Tools: freeGitClaude Code
Agent skills185words6 steps

Wayfinder AI agent planning skill

Orchestrates multi-session project planning into dependency-tracked issue tickets, specs, and prototypes with AI coding agents.

Tools: freeClaude CodeCodex
Agent skills120words4 steps

Claude Code

Claude Code · Cowork17 files

Claude Code Projects

Coordinates multi-thread autonomous coding tasks, pull requests, artifacts, and scheduled routines across repositories.

Tools: freeClaude Code
Claude Code422words12 steps

Verification skill for Claude Code

Automate manual QA checks in Claude Code by codifying verification steps into custom skills.

Tools: freeNodeClaude Code
Claude Code160words4 steps

Claude Code cloud sessions

Run asynchronous coding tasks in isolated cloud environments across desktop, terminal, mobile, and web interfaces.

Tools: freeGitHub accountClaude Code
Claude Code378words8 steps

Write, load and share a Claude Code mod

Build a plugin whose JavaScript or TypeScript hooks watch, rewrite or answer Claude Code events, then validate, load and share it.

Tools: freeClaude Code
Claude Code190words6 steps

Interview, spec and ticket skills for Claude Code

Install a skills plugin, get interviewed before a change, then turn the answers into a spec, tickets or a multi-session map.

Tools: freeClaude Code
Claude Code192words5 steps

Claude Code context and memory habits

Set up CLAUDE.md, trim what fills the context window, give every task a check and reset between tasks.

Tools: freeClaude Code
Claude Code519words9 steps

Spec-Driven Development for AI coding agents

Agree on requirements, implement one feature group, and validate against its specification.

Tools: freeGitClaude Code
Claude Code119words5 steps

MCP servers in Claude Code

Connect an HTTP or local MCP server, choose its scope, and test a tool call.

Tools: freeClaude Code
Claude Code479words3 steps

Writing Claude skills that hold up

Organize skill references and validate the workflow on representative tasks.

Tools: freeClaude Code
Claude Code106words4 steps

Custom subagents in Claude Code FREE SAMPLE

Create a specialist agent with explicit instructions and limited tool access.

Tools: freeClaude Code
Claude Code452words4 steps

Claude Code skills FREE SAMPLE

Create a reusable skill, choose its scope, and test it on a sample task.

Tools: freeClaude Code
Claude Code69words3 steps

Route a coding agent through Free Claude Code

Configure a provider, launch an agent through the proxy, and verify the connection.

Tools: freeAPI keyClaude Code
Claude Code111words4 steps

Build and share Claude skills

Package a tested skill, upload it to Claude, and share the reviewed version.

Tools: freeClaude Code
Claude Code101words4 steps

Claude Code hooks FREE SAMPLE

Configure PreToolUse and PostToolUse hooks in settings.json to auto-format files and block dangerous commands.

Tools: freeClaude Code
Claude Code422words4 steps

Use the Small Business plugin

Install the plugin, connect task inputs, and review its output before taking action.

Tools: freeClaude Code
Claude Code101words4 steps

Claude Code hooks instead of prompt rules

Replace repeatable prompt rules with hooks and test a session-summary trigger.

Tools: freeClaude Code
Claude Code113words4 steps

Unified Claude Chat and Cowork interface

Combines Chat and Cowork modes into a single conversation interface with automatic task routing and output format controls.

Tools: freeAny text chat
Cowork127words3 steps

Local models

Ollama · LM Studio · Model hosting08 files

Ollama Modelfile and companion

A Modelfile that carries your instructions and sets the context window on purpose, with the checks that show it loaded.

Tools: freeOllamaOllama
Ollama573words5 steps

Run a local model with Ollama

Install Ollama, download a model, and test it from the terminal.

Tools: freeOllamaOllama
Ollama90words4 steps

Run Claude Code with a local Ollama model

Connect Claude Code to a local model and verify a small coding task.

Tools: freeOllamaOllama
Ollama446words4 steps

Local AI agent with Ollama and PydanticAI

Deploy local LLMs using Ollama and build tool-calling Python agents with PydanticAI.

Tools: freePythonOllama
Ollama175words6 steps

LM Studio local model runner and server

Download, configure, run, and serve local LLMs via an OpenAI-compatible API to external applications.

Tools: freelm-studioAny AI
LM Studio403words14 steps

A CLAUDE.md for apps with no file system

Put your working rules in the system prompt of Open WebUI, LM Studio or Jan, save them to the preset, and prove they loaded.

Tools: freeOpen WebUILM Studio
LM Studio623words11 steps

Edit files with LM Studio Bionic

Create a project, select a model, and check its file changes.

Tools: freeLM StudioAny AI
LM Studio108words4 steps

Check model hosting constraints before building

Test license fit, memory fit and real inference speed before investing in an app interface.

Tools: freeAny AI
Model hosting130words5 steps

Always-on agents

OpenClaw · Hermes · Pi agent · AI reliability · Agent infrastructure101 files

OpenClaw 2.0 install and setup

Install OpenClaw, configure local and cloud LLMs, create automated agents, and pair with mobile devices.

Tools: freeopenclawAny AI
OpenClaw510words16 steps

OpenClaw setup

Install OpenClaw, connect a model and optional channel, and check its access.

Tools: freeAPI keyAny AI
OpenClaw125words5 steps

Teach Hermes an API once, then enrich a CSV

Teach Hermes Agent an API once to autonomously enrich CSV records using structured web research.

Tools: free alternativeparallel API keyAny AI
Hermes303words3 steps

Lean Hermes skills from repeated tasks

Turn a task you keep repeating or correcting into a tested Hermes skill with a short index entry, on-demand references and helper scripts.

Tools: freeAny AI
Hermes192words6 steps

Run Hermes Agent on a VPS with Telegram

Hermes Agent running as a boot-time service on your VPS, answering on Telegram, remembering preferences and running scheduled jobs.

Tools: freeGitAPI key
Hermes229words7 steps

Add focused tools to an agent workflow

Extract pages, search code, edit prose, or add a coding workflow to a compatible agent.

Tools: freeNodeAny AI
Hermes207words7 steps

Self-compacting extension for the pi coding agent

Builds an extension enabling autonomous coding agents to manage context window limits and compact themselves using custom prompts.

Tools: freeNodejust
Pi agent327words10 steps

Build AI guardrails that actually run

Wire each guard into its real trigger and prove it blocks a known failure.

Tools: freeAny AI
AI reliability122words4 steps

Wire a dataset into an AI-facing data layer

Ship data, connector, vocabulary and consumer together, then verify the served result.

Tools: freeAny AI
AI reliability116words5 steps

Verify a SQL View with an Independent Oracle

Compare SQL-derived tier metrics with a separate implementation over the same raw rows.

Tools: freeAny AI
AI reliability123words5 steps

Verify inherited claims at the owning source

Treat plans and summaries as hypotheses; check each action-driving claim against the source that owns it.

Tools: freeAny AI
AI reliability117words4 steps

Prove which inference route actually executes

Trace selected branches instead of treating zero cost as proof of a particular provider.

Tools: freeAny AI
AI reliability91words5 steps

Verify flight recovery after command loss

Expire stale commands independently and confirm recovery through actual mode telemetry.

Tools: freeAny AI
AI reliability139words5 steps

Verify writes in the rendered view

Confirm a write through the user-visible record, not its immediate API response.

Tools: freeAny AI
AI reliability144words5 steps

Test Separate Entry and Protection in a Simulator

Exercise delayed protection, submission failures, and restart reconciliation using synthetic orders.

Tools: freeAny AI
AI reliability169words6 steps

Validate Text Against Explicit Constraints

Check output limits and forbidden text separately, then verify that the response still fulfills its task.

Tools: freeAny AI
AI reliability134words5 steps

Preserve experiment runs and count every trial

Keep immutable run artifacts and a complete variant ledger before selecting results.

Tools: freeAny AI
AI reliability138words4 steps

Validate AI Content Before Committing

Accept generated document content only after schema validation, while preserving existing styling.

Tools: freeAny AI
AI reliability133words5 steps

Test verification ladders without network calls

Inject one lookup seam to exercise source agreement, fallback, errors, and freshness boundaries deterministically.

Tools: freeAny AI
AI reliability161words6 steps

Validate Locking Without Launching Real Workers

Exercise contention, process detection, and clone isolation with temporary locks and fake launchers.

Tools: freeAny AI
AI reliability164words6 steps

Test Geographic Coverage Invariants

Separate stable synthetic grain tests from a live-data smoke test that respects the service footprint.

Tools: freeAny AI
AI reliability155words7 steps

Test a UI control by its downstream effect

Catch a picker that passes values through several layers but changes nothing.

Tools: freeAny AI
AI reliability95words5 steps

Expose retrieval failures through fallback

Keep retrieval infrastructure errors distinguishable from valid searches with no answer.

Tools: freeAny AI
AI reliability182words6 steps

Keep live AI updates from overwriting user edits

Validate streaming updates against the whole document and let local edits win through completion.

Tools: freeAny AI
AI reliability132words5 steps

Scope capabilities to each pipeline stage

Give fetch stages spending access and notification stages only their own send capability.

Tools: freeAny AI
AI reliability120words5 steps

Separate verified facts from inference

Give exact sourced claims and useful hypotheses separate generation and validation rules.

Tools: freeAny AI
AI reliability176words6 steps

Prevent an action from approving its own plan

Make approval a separate recorded transition and require the simulated submission to match the approved terms.

Tools: freeAny AI
AI reliability115words4 steps

Scope a recurring defect before fixing it

Find every implementation of the failing operation before changing the first example.

Tools: freeAny AI
AI reliability96words6 steps

Verify revocation stops active media

Authorize each media session and test revocation while a viewer is connected.

Tools: freeAny AI
AI reliability138words5 steps

Restore global module mocks after each test file

Prevent test order from turning an isolated pass into a full-suite failure.

Tools: freeAny AI
AI reliability109words6 steps

Resolve the subject once for every build path

Use one record resolver across entry points and preserve an unresolved subject instead of substituting a nearby match.

Tools: freeAny AI
AI reliability106words4 steps

Demote Duplicate Signals to Cross-Checks

Stop a summary engine from voting again on concepts already covered by dedicated inputs.

Tools: freeAny AI
AI reliability119words4 steps

Record Deliberation Confidence Changes

Preserve independent initial confidence and final confidence to measure what deliberation actually changes.

Tools: freeAny AI
AI reliability149words5 steps

Prevent lookahead in historical backtests

Limit each signal to information available at its decision time.

Tools: freeAny AI
AI reliability122words4 steps

Retire legacy engines while protecting customers

Freeze customer-bearing units while replacing the engine; archive obsolete code and docs with recoverable history.

Tools: freeAny AI
AI reliability137words6 steps

Keep Incompatible Signal Baselines Separate

Keep signals separate when their questions and comparison baselines are incompatible.

Tools: freeAny AI
AI reliability127words5 steps

Prepare a Voice-Cloning Reference

Control reference length, language, and voice instructions before judging cloned speech.

Tools: freeAny AI
AI reliability135words5 steps

Check evidence before spending writer calls

Validate adapter evidence before generation and bound repeated rejected work across scheduled runs.

Tools: freeAny AI
AI reliability151words5 steps

Normalize and Validate Structured Model Responses

Extract answer text from mixed content blocks before validating structured output.

Tools: freeAny AI
AI reliability111words4 steps

Register computed facts as narrator anchors

Give the writer and claim validator the same approved facts instead of hiding facts in framing.

Tools: freeAny AI
AI reliability109words5 steps

Match forbidden labels without substring false alarms

Check rendered output for prohibited labels while allowing unrelated words.

Tools: freeAny AI
AI reliability131words4 steps

Lint causal claims between agent outputs

Flag causal connectors joining different agent references while allowing explicit source attribution.

Tools: freeAny AI
AI reliability123words5 steps

Propagate freshness and trust through confidence

Compute confidence from source quality and upstream dependencies while exposing its limits.

Tools: freeAny AI
AI reliability135words4 steps

Keep an explainer chat honest without live data

Constrain an ungrounded chat to capabilities and direct precise data requests to a grounded workflow.

Tools: freeAny AI
AI reliability173words6 steps

Make repaired migration exemptions fail visibly

Use inverted assertions so a fixed exception cannot remain silently exempt.

Tools: freeAny AI
AI reliability100words5 steps

Show an honest degraded state when feeds fail

Define a floor below which an empty-tolerant page must report unavailable data.

Tools: freeAny AI
AI reliability124words5 steps

Prove checks with deliberate failure fixtures

A detector passes only after catching a known failure and accepting a valid case.

Tools: freeAny AI
AI reliability118words4 steps

Ground charts and answers on one data artifact

Publish deterministic, grain-specific data once, then use it for charts and grounded answers.

Tools: freeAny AI
AI reliability238words6 steps

Rebuild frozen deliverables after generation changes

Detect invented figures, minted links, missing disclosures and caption-limit regressions on fixed inputs.

Tools: freeAny AI
AI reliability93words4 steps

Generate Tool Traces Through a Verifier Gate

Separate tool selection from prediction generation and retain rejected traces without admitting them to accepted data.

Tools: freeAny AI
AI reliability172words7 steps

Gate Records with Pure, Versioned Verifiers

Admit only passing candidates, then independently reverify saved records with immutable offline rules.

Tools: freeAny AI
AI reliability150words6 steps

Gate and Route Pasted Listing Links

Default fetching off, route link requests exclusively, and request missing facts instead of guessing.

Tools: freeAny AI
AI reliability262words6 steps

Separate sunk spending from a resource decision record

Prepare a forward-looking resource review while preserving relevant costs and history.

Tools: freeAny AI
AI reliability146words4 steps

Flag Retrospective Forecast Contamination

Keep retrospective questions inside a model's training window from being counted as clean forecast tests.

Tools: freeAny AI
AI reliability143words5 steps

Test kill-switch latch creation in isolation

Exercise the enforcement path in temporary state and confirm a simulated breach creates its latch.

Tools: freeAny AI
AI reliability122words4 steps

Review Untrusted Posts Without Executing Them

Keep scraped instructions inert, label evidence quality, and escalate consequential recommendations.

Tools: freeAny AI
AI reliability162words5 steps

Enforce guards in the action path

Put each critical guard in the module that performs the action, then verify the submit path enforces it.

Tools: freeAny AI
AI reliability132words4 steps

Emit Normalized Trial Facts

Convert trial evidence into typed contexts and stable fields that batch reports can aggregate.

Tools: freeAny AI
AI reliability144words5 steps

Expose rejected prose in build reports

Make silent narrative drops visible without weakening the claim gate.

Tools: freeAny AI
AI reliability128words4 steps

Edit generated content through a gated rebuild

Change structured inputs and regenerate through the original gates instead of editing generated prose directly.

Tools: freeAny AI
AI reliability131words5 steps

Exercise kill conditions with seeded cases

Prove breach detection and alert delivery using isolated synthetic fixtures.

Tools: freeAny AI
AI reliability127words5 steps

Draft Evidence-Grounded Answer Keys

Define required facts, connections and reasoning before grading a scenario response.

Tools: freeAny AI
AI reliability163words6 steps

Make research discoverable to coding agents

Give agents a searchable route to research that ordinary repository searches miss.

Tools: freeAny AI
AI reliability124words4 steps

Audit model additions before composing factual prose

Allow code-written facts while rejecting digits introduced by the model.

Tools: freeAny AI
AI reliability132words5 steps

Diagnose Generators with Verifier Kill Reports

Compare identical task corpora and isolate semantic failures hidden by valid schemas and tool calls.

Tools: freeAny AI
AI reliability166words7 steps

Choose guard types before building a feature

Pair concrete failure modes with controls that can observe them during design.

Tools: freeAny AI
AI reliability134words4 steps

Design an agent workflow from its target answer

Identify the questions, evidence and conflict rules needed before building one useful slice.

Tools: freeAny AI
AI reliability127words6 steps

Keep Agent Actions Within Their Deadline

Limit each model call and switch to a deterministic policy when the response cannot supply usable actions.

Tools: freeAny AI
AI reliability153words6 steps

Include pending orders in simulation limits

Count unfilled commitments alongside simulated holdings so repeated submissions cannot bypass exposure limits.

Tools: freeAny AI
AI reliability117words4 steps

Compute claims before asking AI for prose

Give the narrator settled facts and omit any paragraph that fails its claim gate.

Tools: freeAny AI
AI reliability113words4 steps

Compare Speech Quality Controls

Use controlled listening comparisons to select speech settings and measure their runtime cost.

Tools: freeAny AI
AI reliability155words5 steps

Compare Distillations on Identical Evidence

Evaluate wording, factual fidelity, classifications, and dropped spans without overstating recall.

Tools: freeAny AI
AI reliability160words5 steps

Classify billing refusals at the call boundary

Route a confirmed billing refusal through the existing incident path without retrying it.

Tools: freeAny AI
AI reliability133words5 steps

Cap rule-guided synthesis revision

Revise a failing synthesis twice, then expose unresolved violations for human review.

Tools: freeAny AI
AI reliability93words5 steps

Check builder repeatability across fresh contexts

Compare repeated builds while isolating only declared nondeterministic fields.

Tools: freeAny AI
AI reliability125words5 steps

Keep internal errors out of API responses

Return a generic error while retaining diagnostic detail in protected logs.

Tools: freeAny AI
AI reliability119words4 steps

Stop AI reports inventing relationships

Audit comparisons, trends, counts and sequences even when every number is real.

Tools: freeAny AI
AI reliability120words5 steps

Match verification evidence to bug behavior

Choose a real observation that can expose the failure being fixed.

Tools: freeAny AI
AI reliability135words4 steps

Approve the exact reviewed page version

Approve only when the current page matches the reviewed version and the brief actually displayed.

Tools: freeAny AI
AI reliability163words6 steps

Apply domain constraints before synthesis

Evaluate binding domain rules before accepting a synthesized recommendation.

Tools: freeAny AI
AI reliability112words5 steps

Verify Protocol Routes With POST Discovery and Calls

Catch mount errors that GET probes miss by exercising discovery and a real tool call.

Tools: freeAny AI
Agent infrastructure112words4 steps

Separate Drone Control from Video

Keep flight authority separate from observation, with recovery independent of the hub.

Tools: freeAny AI
Agent infrastructure105words5 steps

Share classification across warning and stop gates

Use one classifier so advance warnings and final enforcement cannot disagree.

Tools: freeAny AI
Agent infrastructure75words5 steps

Scope Obstacle Checks to Each Command Path

Validate obstacle coverage and sensor-loss behavior for the modes and commands actually used.

Tools: freeAny AI
Agent infrastructure128words5 steps

Route tool calls by supported capabilities

Match questions against available evidence and coverage instead of broad topic keywords.

Tools: freeAny AI
Agent infrastructure101words5 steps

Restrict downstream agents to validated outputs

Expose compact upstream results through explicit dependency edges and reject incomplete contracts.

Tools: freeAny AI
Agent infrastructure111words5 steps

Require read evidence from the editing session

Keep another session's recent read from satisfying an editor's preparation gate.

Tools: freeAny AI
Agent infrastructure84words5 steps

Register plan scope when a spec is saved

Capture scope after spec writes, merge it into the session plan, and report overlaps.

Tools: freeAny AI
Agent infrastructure212words6 steps

Prove Aircraft Interfaces Before Selection

Bench-prove commands, telemetry and recorder-bound video for the exact aircraft configuration.

Tools: freeAny AI
Agent infrastructure127words5 steps

Submit Orders Once Per Observed Turn

Use visible state, validate per-unit actions, and recover from stale turns without duplicating accepted submissions.

Tools: freeAny AI
Agent infrastructure203words7 steps

Serialize Runners Across Isolated Worktrees

Use one account-wide lock and detect processes that launched outside it.

Tools: freeAny AI
Agent infrastructure197words6 steps

Isolate Sessions Before Integrating Changes

Give each session its own working tree and integrate committed changes through fast-forward updates.

Tools: freeGitAny AI
Agent infrastructure152words5 steps

Monitor a home agent server

Separate host resource visibility, container metrics and proven alert delivery.

Tools: freeAny AI
Agent infrastructure126words5 steps

Require a durable session log before pushing

Preserve session history and block pushes that omit its update.

Tools: freeAny AI
Agent infrastructure115words4 steps

Distinguish a failed agent from a quiet agent

Monitor failed runs, missed runs and invalid outputs as three separate failure paths.

Tools: freeAny AI
Agent infrastructure133words5 steps

Expose failures in required rule loading

Prevent missing, blank or unreadable rule files from silently replacing the active rule set with stale defaults.

Tools: freeAny AI
Agent infrastructure93words4 steps

Evaluate Shared-Repo Agent Coordination

Test overlapping plans, exclusive file access, automatic recovery, queue fairness, and coordination context.

Tools: freeAny AI
Agent infrastructure316words9 steps

Coordinate parallel agents editing shared files

Compare planned scopes, gate conflicting writes, and release claims using version-control evidence.

Tools: freeAny AI
Agent infrastructure182words6 steps

Claim a Handoff Before Launching

Give each handoff a visible owner so duplicate recipients cannot independently launch the same work.

Tools: freeAny AI
Agent infrastructure118words4 steps

Carry answer context in every tool payload

Deliver the synthesis dossier and compact rules together so follow-ups can reuse retrieved context.

Tools: freeAny AI
Agent infrastructure113words4 steps

Capture corrections before answering

Persist user issues immediately and carry unresolved items across sessions.

Tools: freeAny AI
Agent infrastructure83words3 steps

OpenAI & editors

Codex · ChatGPT agents · Cursor · Antigravity08 files

Codex Ultrafast mode on the Pro 500 plan

Configure Ultrafast mode in Codex to rapidly generate interactive WebGL, 3D, and full-stack web applications.

Tools: paidCodex
Codex272words5 steps

Six-step method for Codex skills

Create, test, verify, and optimize markdown-based skills for AI agents.

Tools: freeCodexClaude Code
Codex342words6 steps

Navigate Codex CLI sessions

Fork a conversation, inspect agent threads, and check session usage.

Tools: freeCodex
Codex89words4 steps

Start a project with Codex

Open a local project, plan a change, and review the implementation.

Tools: freeOpenAI accountCodex
Codex101words4 steps

ChatGPT Browser and Computer Use

Automate research, web workflows via Chrome, and local desktop applications using natural-language instructions.

Tools: freeOpenAI accountAny text chat
ChatGPT agents261words4 steps

Cursor Projects agent coordinator

Coordinates autonomous multi-agent coding tasks across cloud and local machines with persistent threads and event triggers.

Tools: freeCursor
Cursor113words0 steps

Google Antigravity 2.0 agent platform

Dispatches and manages parallel AI agents to plan, code, debug, and run automated tasks.

Tools: freePythonAny AI
Antigravity206words6 steps

Antigravity Remote Control

Monitor and steer coding agents across machines from a mobile device without syncing local build environments.

Tools: freeGoogle accountAny AI
Antigravity124words4 steps

Automation & repos

n8n · Browser agents · Computer skills · GitHub repos · App architecture · Monitoring products64 files

n8n Agents tab (early preview)

Build, configure, and deploy standalone or workflow-integrated AI agents natively inside n8n without complex node chains.

Tools: freen8nAPI key
n8n334words11 steps

Model benchmark for n8n workflow generation

Compares six frontier AI models on cost, speed, and reliability when generating complex n8n workflows via MCP.

Tools: freen8nAPI key
n8n349words7 steps

n8n Telegram agent

Build and connect an automated personal AI assistant to Telegram using n8n Agents, MCP tools, and skills.

Tools: freen8n accountTelegram account
n8n668words21 steps

n8n MCP and skills for Claude Code

Install n8n skills and configure MCP in Claude Code to autonomously generate, configure, and test n8n workflows.

Tools: freen8nClaude Code
n8n374words13 steps

n8n instance-level MCP setup

Connect AI coding agents to an n8n instance to generate and edit workflows automatically.

Tools: freen8nClaude Code
n8n464words20 steps

Connect n8n agents and workflows

Build an agent, expose a workflow tool, and call the agent from another workflow.

Tools: freeAPI keyAny AI
n8n146words5 steps

Your first AI agent in n8n

Build Q&A ingestion and retrieval workflows, then test answers against stored records.

Tools: freeAPI keyAny AI
n8n186words5 steps

Model Context Protocol for Playwright QA

Connects AI agents to testing tools, automation, and databases using Model Context Protocol (MCP) and Playwright.

Tools: freeNodePython
Browser agents231words6 steps

Claude Code Chrome extension setup

Connect Claude Code to Google Chrome to automate browser navigation, capture screenshots, and inspect web pages.

Tools: freeClaude Code
Browser agents228words4 steps

Linux command-line troubleshooting tools

Run Linux commands inside Docker to inspect logs, find files, check ports, and edit configuration files.

Tools: freeDockerAny AI
Computer skills369words19 steps

AI web scraping and classification pipeline

Scrape dynamic web pages using Playwright and Claude Code, route through proxies, and classify text using Jev.

Tools: freePythonNode
Computer skills385words7 steps

Isolate IoT devices on an OPNsense VLAN

An IoT VLAN on OPNsense with its own DHCP range and internet access, blocked from your trusted networks.

Tools: freeAny AI
Computer skills205words7 steps

Preserve the original command's exit status

Read saved output after capturing the producer's status, not through a masking pipeline.

Tools: freeAny AI
Computer skills82words5 steps

Avoid quoting bugs across Windows and Linux

Put complex commands in scripts and verify values after crossing a shell boundary.

Tools: freeAny AI
Computer skills127words5 steps

Test redaction with a known matching fixture

Confirm a redactor changes fake credentials instead of merely compiling successfully.

Tools: freeAny AI
Computer skills122words5 steps

drive-screen coding agent skill

Drives desktop GUIs using CLI Python scripts executed directly by terminal coding agents without dedicated computer-use harnesses.

Tools: freePythonAny AI
Computer skills157words6 steps

Add testing and design tools to a coding agent

Install selected tools for UI checks, skills evaluation, and code linting.

Tools: freeNodeAny AI
GitHub repos246words8 steps

Install tools for an AI development workflow

Set up a selected converter, terminal, review tool, gateway, or diagramming skill.

Tools: freeNodePython
GitHub repos223words9 steps

Map an app workflow to running components

Give every step a runtime, owner, state contract and honest implementation status.

Tools: freeAny AI
App architecture122words5 steps

Verify a Vendor Contract Before Coding

Use live documentation and one real response to establish the API contract before implementing an integration.

Tools: freeAny AI
App architecture138words6 steps

Typecheck scripts excluded from the app build

Give scripts an explicit compiler project instead of trusting inferred editor diagnostics.

Tools: freeAny AI
App architecture96words6 steps

Triage routes for safe revalidation

Match public-route caching to source cadence and measure actual egress before and after.

Tools: freeAny AI
App architecture144words5 steps

Use one rule to surface work needing attention

Share attention logic across command and session views so draft and clockless work cannot disappear.

Tools: freeAny AI
App architecture116words5 steps

Keep lookup links separate from source citations

Make feed items navigable without turning convenience links into evidence in exported notes.

Tools: freeAny AI
App architecture155words4 steps

Score Outcomes Independently of Publication

Keep outcome grading running when optional feeds or narrative publication fail.

Tools: freeAny AI
App architecture157words6 steps

Register validators by content shape

Let surfaces share validation by declaring the kind of content they produce.

Tools: freeAny AI
App architecture81words5 steps

Route builds by explicit tags across entry points

Keep editable prompt text separate from the identity that selects a builder.

Tools: freeAny AI
App architecture121words5 steps

Roll out generated content with a cold fallback

Approve a single-entity pilot and batch previews before switching readers to new structured content.

Tools: freeAny AI
App architecture184words6 steps

Retry transient profile failures before degrading

Treat confirmed signed-out state differently from a temporary profile-fetch failure.

Tools: freeAny AI
App architecture93words5 steps

Render Saved Charts as Share Cards

Generate consistent PNG cards while preserving chart values and source provenance.

Tools: freeAny AI
App architecture172words5 steps

Render answers by question complexity

Keep everyday answers conversational and reserve full audit structure for explicit audit requests.

Tools: freeAny AI
App architecture127words6 steps

Detect redirects that shadow real pages

Catch a deployed route that cannot render because a redirect wins first.

Tools: freeAny AI
App architecture118words6 steps

Serve Files With Expiring Private Tokens

Protect downloadable repairs with loopback serving, per-request hashes, and bounded secret-token lifetimes.

Tools: freeAny AI
App architecture202words7 steps

Guard Only Corridor-Dependent Lookups

Use the resolved county to retain applicable results and skip only the lookup requiring a corridor ID.

Tools: freeAny AI
App architecture125words5 steps

Package runtime assets with their routes

Include native dependencies and disk-read assets before shipping server-rendered outputs.

Tools: freeAny AI
App architecture132words5 steps

Isolate SVG IDs for Repeated Rendering

Keep gradients, masks, clips, and local references attached to the correct icon instance.

Tools: freeAny AI
App architecture166words5 steps

Isolate Dataset Load and Parse Failures

Keep missing or malformed datasets from aborting an otherwise usable multi-dataset response.

Tools: freeAny AI
App architecture123words5 steps

Gate Actions by Metered Usage

Enforce weekly action limits with configuration that defaults to unlimited access.

Tools: freeAny AI
App architecture187words6 steps

Fork content edits to preserve shared snapshots

Create a new version for refreshed data or content changes while keeping the shared original frozen.

Tools: freeAny AI
App architecture149words6 steps

Stop failed reads from creating duplicate projects

Separate a confirmed empty project list from a failed lookup before automatic creation.

Tools: freeAny AI
App architecture97words6 steps

Consolidate Database Client Imports

Remove a competing client configuration that can silently break authentication.

Tools: freeAny AI
App architecture97words5 steps

Bridge a proven renderer into a new project

Transfer an established rendering capability through a small specification contract.

Tools: freeAny AI
App architecture113words5 steps

Prove the core flow before adding product gates

Drive the real user journey before designing pricing, throttling, or additional restrictions.

Tools: freeAny AI
App architecture86words5 steps

Gate Paid Exports with User-Scoped Reads

Separate privileged billing lookup from user-scoped export reads and define billing-outage behavior.

Tools: freeAny AI
App architecture176words6 steps

Confirm and revoke a scoped change decision

Tie a short-lived approval to a verified identity and recheck the target before confirmation.

Tools: freeAny AI
App architecture160words5 steps

Bind Page Approval to Content Changes

Invalidate approval when meaningful page content changes while allowing freshness metadata to update.

Tools: freeAny AI
App architecture141words5 steps

Validate Feed Shapes Before Trusting Zero

Inspect live responses and repair parsing before trusting an empty discovery feed.

Tools: freeAny AI
Monitoring products152words5 steps

Trigger Brief Review on Title Changes

Detect headline corrections under stable IDs without reviewing every rotating search excerpt.

Tools: freeAny AI
Monitoring products175words4 steps

Run a Source-Backed Monitoring Loop

Deduplicate collected items, rewrite a linked brief when new information arrives, and notify only on a brief change.

Tools: freeAny AI
Monitoring products191words7 steps

Build a monitor from probe to dashboard

Normalize live records, preserve a baseline, detect changes and render a finished dashboard.

Tools: freeAny AI
Monitoring products126words5 steps

Measure Operator Review Load During a Pilot

Record reviews and minutes per subject over seven days.

Tools: freeAny AI
Monitoring products121words5 steps

Retain and Recheck Discrepancies

Preserve unresolved observations and notify users when comparable observed states change.

Tools: freeAny AI
Monitoring products159words6 steps

Add time bounds to aggregate usage counters

Keep a counter capable of answering when activity began and last occurred.

Tools: freeAny AI
Monitoring products86words5 steps

Provision a Tracked Subject from a Request

Map a request to probed sources, a persistent monitoring unit, and its requester-facing page.

Tools: freeAny AI
Monitoring products167words6 steps

Prove the Monitoring Customer Journey

Test request intake, published results, delivered changes, quiet repeats, and notification controls before charging.

Tools: freeAny AI
Monitoring products148words7 steps

Archive Retired Code Without Breaking Active Imports

Trace active and frozen consumers before moving retired components out of current discovery.

Tools: freeAny AI
Monitoring products149words6 steps

Send notifications only on real changes

Write the dashboard first, then notify on meaningful diffs with honest delivery outcomes.

Tools: freeAny AI
Monitoring products113words5 steps

Ingest Tracked Events Idempotently

Merge repeated event IDs and distinguish stale observation from an empty scene.

Tools: freeAny AI
Monitoring products133words5 steps

Generate watcher coverage from workflow files

Derive watcher subscriptions and monitoring metadata from one generated workflow manifest.

Tools: freeAny AI
Monitoring products116words5 steps

Build public examples from real product output

Generate example cards and pages from sourced, committed output rather than hand-entered marketing figures.

Tools: freeAny AI
Monitoring products112words5 steps

Add customers through configuration

Keep one probe, diff and rendering engine while each customer owns settings and history.

Tools: freeAny AI
Monitoring products117words5 steps

Consolidate pipeline health reporting

Combine existing health checks into one machine report that backs the operator inventory.

Tools: freeAny AI
Monitoring products157words6 steps

Compare Observed Offer State

Compare catalog prices, processed listings, and variant-specific page evidence without treating uncertainty as healthy.

Tools: freeAny AI
Monitoring products169words5 steps

Test data availability before accepting work

Run the real extraction against the real target and issue an evidence-based viability receipt.

Tools: freeAny AI
Monitoring products121words5 steps

Data & workflows

Data engineering · Email production · Operations213 files

Compute ZIP Rent Yields With Sparse Flood Data

Combine value, monthly rent, and optional flood adjustments while preserving missing and implausible results.

Tools: freeAny AI
Data engineering235words5 steps

Define Export Fields in One Registry

Drive every export through an ordered public-field schema that excludes internal data.

Tools: freeAny AI
Data engineering130words5 steps

Verify Identity Before Comparing Records

Activate only verified deterministic mappings and keep unmatched records visible in coverage.

Tools: freeAny AI
Data engineering129words5 steps

Keep Calendar Buckets Independent of Session Timezones

Check date types and pin timestamp bucketing to a declared calendar timezone.

Tools: freeAny AI
Data engineering130words5 steps

Detect Stale Price Feeds

Inspect record timestamps and collector status before accepting a feed as current.

Tools: freeAny AI
Data engineering144words5 steps

Validate Browser Date Filters

Detect silently rejected date values before trusting search results.

Tools: freeAny AI
Data engineering122words5 steps

Centralize exact normalized street comparison

Handle compound-word spacing consistently without allowing partial matches.

Tools: freeAny AI
Data engineering93words5 steps

Build latest snapshots with calendar change rates

Calculate monthly and annual changes using the newest observation within a seven-day calendar tolerance.

Tools: freeAny AI
Data engineering181words6 steps

Verify a Listing Link Selects the Correct Variant

Repair an exact variant-link mismatch and verify the destination by fetching it.

Tools: freeAny AI
Data engineering111words5 steps

Repair a Stale Override at Its Owning Source

Test simultaneous source precedence and preserve unrelated fields through the next synchronization.

Tools: freeAny AI
Data engineering114words5 steps

Validate Sale Timing Across Both Boundaries

Correct an established sale interval and observe effective prices before, during, and after it.

Tools: freeAny AI
Data engineering111words5 steps

Repair Listing Availability From Sellability

Align a listing with established storefront sellability while leaving physical inventory unchanged.

Tools: freeAny AI
Data engineering110words5 steps

Fill Missing Identifiers From Verified Catalog Data

Transfer an authentic identifier already held for the exact variant; refuse invented replacements.

Tools: freeAny AI
Data engineering109words4 steps

Suppress Thin-Sample Local Sale Medians

Calculate local housing-price medians with a labeled county fallback for groups containing fewer than 20 sales.

Tools: freeAny AI
Data engineering188words5 steps

Correct unit errors with superseding records

Fix ambiguous units without erasing the original prediction or its eventual outcome.

Tools: freeAny AI
Data engineering114words5 steps

Build a staged, rerunnable data refinery

Preserve raw records, filter and synthesize relevant facts, then validate output before replacing an existing artifact.

Tools: freeAny AI
Data engineering155words7 steps

Harden ingestion with keys and contracts

Choose merge or replacement from observed key behavior and guard required fields.

Tools: freeAny AI
Data engineering112words5 steps

Name computed quantities from source semantics

Check what a field measures before attaching a stronger label to correct arithmetic.

Tools: freeAny AI
Data engineering119words5 steps

Separate source capability from ingested data

Record what a source exposes alongside the subset your pipeline actually stores.

Tools: freeAny AI
Data engineering108words5 steps

Anchor report values to a source snapshot

Give every decision-facing value an observation time and an identifiable snapshot.

Tools: freeAny AI
Data engineering78words4 steps

Deliver and Verify Spreadsheet CSV Files

Preserve Unicode, serialize every cell, and verify downloaded values in the target spreadsheet.

Tools: freeAny AI
Data engineering151words5 steps

Separate Trial Facts From Batch Analysis

Aggregate normalized facts automatically and configure semantic analysis separately.

Tools: freeAny AI
Data engineering131words5 steps

Keep nearby and resident amenities distinct

Preserve location grain and uncertainty instead of turning proximity into access.

Tools: freeAny AI
Data engineering136words6 steps

Keep capped histories from producing false totals

Expose truncation and withhold totals across unrelated record cycles.

Tools: freeAny AI
Data engineering97words6 steps

Scope Reconciliation Checks by Grain and Window

Validate signed summary totals only within matching scope; keep that check separate from detail-row verification.

Tools: freeAny AI
Data engineering146words6 steps

Scope an Ingestion Source Before Loading

Inventory the source schema, choose its canonical root, and connect the consuming product before ingesting.

Tools: freeAny AI
Data engineering205words7 steps

Run opt-in value and schema quality probes

Configure table-level value checks and compare live column types with reviewed baselines.

Tools: freeAny AI
Data engineering194words7 steps

Retain raw API responses for later parsing

Persist successful source responses before deriving downstream fields.

Tools: freeAny AI
Data engineering177words6 steps

Parse typed families from stored response envelopes

Recover one-to-many data from retained payloads without buying the same response again.

Tools: freeAny AI
Data engineering166words5 steps

Reject Ambiguous Dates at Rule Boundaries

Withhold partial dates when their possible values cross different rule windows.

Tools: freeAny AI
Data engineering152words6 steps

Reconcile Primary Facts with Secondary Crosschecks

Reconcile source coverage and surface proxy disagreements without letting them overwrite primary facts.

Tools: freeAny AI
Data engineering160words7 steps

Reconcile Coverage Against Record Geography

Separate missing geographic mappings from out-of-footprint records before widening dataset coverage.

Tools: freeAny AI
Data engineering157words6 steps

Make serving views read one typed data root

Remove duplicate parsing while preserving consumer-facing validation.

Tools: freeAny AI
Data engineering96words5 steps

Ratify Data Authorities by Purpose

Make each source recommendation independently approvable for the question it answers.

Tools: freeAny AI
Data engineering135words5 steps

Parse Ranges Without Hiding Missing Indicators

Preserve range endpoints, support alternate phrasing, and fail ingestion when any required indicator produces no rows.

Tools: freeAny AI
Data engineering142words6 steps

Check active ingest destinations after changes

Detect registered pipelines whose claimed destination was never created.

Tools: freeAny AI
Data engineering131words4 steps

Probe data using the consumer's query shape

Keep an expensive count or wrong schema from creating a false outage diagnosis.

Tools: freeAny AI
Data engineering109words5 steps

Probe real records before building a pipeline

Check keys, fields, pagination and runtime on a small pull before committing to ingestion.

Tools: freeAny AI
Data engineering119words5 steps

Preserve Unknowns in Verified Snapshots

Require evidence for negative findings and keep interpreted document fields outside verified facts.

Tools: freeAny AI
Data engineering127words6 steps

Preserve Observation Meaning and Vintages

Keep measurement identity, missingness, availability, and revisions explicit.

Tools: freeAny AI
Data engineering165words6 steps

Separate Valuation Sources and Labels

Keep assessed values, asking prices, recorded sales, and value indexes in their proper roles.

Tools: freeAny AI
Data engineering118words5 steps

Preserve enriched fields during routine upserts

Keep a sparse recurring feed from erasing values supplied by enrichment.

Tools: freeAny AI
Data engineering112words5 steps

Preserve source date precision through rendering

Use the same grain-aware formatter for visible dates and permitted model numerals.

Tools: freeAny AI
Data engineering117words5 steps

Prepare Survey Microdata for Calibration

Select appropriate person-level data, check coverage, extract observed values, and derive complete calibration margins.

Tools: freeAny AI
Data engineering252words7 steps

Parse signed percentages without hiding conflicts

Accept signed percentages, reject currency-shaped strings, and expose contradictory signs.

Tools: freeAny AI
Data engineering89words5 steps

Gate a data-path cutover on output parity

Compare numeric, aggregate, ranked-output, and table parity across three rebuilds, then prove mismatch detection.

Tools: freeAny AI
Data engineering275words8 steps

Keep repeated scrapes from renewing old news

Rank content by publication or preserved first-seen time instead of its latest scrape.

Tools: freeAny AI
Data engineering79words5 steps

Register a Corridor Across Profile and City Registries

Keep corridor identity, alias resolution, city membership, and counts consistent during onboarding.

Tools: freeAny AI
Data engineering171words6 steps

Guard Fraction and Percentage Swaps

Prevent a source replacement from silently changing a share by a factor of one hundred.

Tools: freeAny AI
Data engineering133words5 steps

Convert Listing Totals to Unit Prices

Establish whether a price covers one item or the entire listing before comparing values.

Tools: freeAny AI
Data engineering146words4 steps

Report Pipeline Freshness From a Cadence Registry

Use registered cadences and actual load metadata to publish a nonblocking stale-pipeline summary.

Tools: freeAny AI
Data engineering142words5 steps

Mark clipped source lists as incomplete

Prevent a truncated entity list from being consumed as the complete record.

Tools: freeAny AI
Data engineering88words5 steps

Map a dataset to a constrained schema

Separate observed mappings from grounded inference, normalize values, and validate output.

Tools: freeAny AI
Data engineering235words6 steps

Make a backfill resumable per record

Use missing-field selection and stable record keys instead of restarting a sampling grid.

Tools: freeAny AI
Data engineering114words6 steps

Log generated predictions and linked outcomes

Persist every synthesis run and join later observations to its original prediction.

Tools: freeAny AI
Data engineering119words5 steps

Size History Windows for Reporting Delays

Retrieve enough calendar history to retain the usable observations required by a rolling calculation.

Tools: freeAny AI
Data engineering130words5 steps

Carry geographic grain with computed values

Compute wider statistics from underlying records and keep their scope attached.

Tools: freeAny AI
Data engineering89words5 steps

Inventory a Records Search Output Schema

Capture every result field and unresolved interpretation before designing ingestion.

Tools: freeAny AI
Data engineering127words6 steps

Implement a Deterministic ZIP Resolver

Resolve scope and county from static data, preserving unknown classifications and stable place choices.

Tools: freeAny AI
Data engineering146words5 steps

Quarantine Suspect Extraction Rows

Block validation failures from downstream use until a reviewer explicitly resolves them.

Tools: freeAny AI
Data engineering111words5 steps

Hold Data With Conflicting Measure Definitions

Block release when accurate values may represent a different measure from the one the destination consumes.

Tools: freeAny AI
Data engineering133words5 steps

Hand Off Immutable Observation Exports

Exchange read-only observation files with a manifest instead of sharing a mutable database.

Tools: freeAny AI
Data engineering156words6 steps

Exclude Rollups from Local Aggregates

Prevent regional and intermediate totals from entering local selections or sums.

Tools: freeAny AI
Data engineering119words4 steps

Check Geographic Coverage of Latest Views

Expose a latest-record view that silently drops part of its intended service area.

Tools: freeAny AI
Data engineering114words4 steps

Guard destructive refreshes before writing

Keep good data when a replacement pull is empty, incomplete or missing required fields.

Tools: freeAny AI
Data engineering108words5 steps

Resolve geographic entities with geometry

Use authoritative boundaries with coordinate and ambiguity checks when names do not encode membership.

Tools: freeAny AI
Data engineering112words5 steps

Generate Validated Active Flags

Convert supported research into at most five schema-valid event flags.

Tools: freeAny AI
Data engineering134words5 steps

Require complete coverage before reporting zero

Preserve held data during outages and prevent missing collection windows from becoming invented zeros.

Tools: freeAny AI
Data engineering135words5 steps

Gate History in a Ragged Time-Series Panel

Determine eligibility per series and count usable series-days instead of quoting the nominal date span.

Tools: freeAny AI
Data engineering146words6 steps

Require Schedules and Dry Runs With New Pipelines

Generate pipeline companions together and block changes that omit scheduling or dry-run support.

Tools: freeAny AI
Data engineering101words5 steps

Gate Retirement on Replacement Readiness

Keep overlapping datasets until the replacement exists and approved consumers have moved.

Tools: freeAny AI
Data engineering139words5 steps

Gate batches with content contracts

Quarantine invalid rows before merge; abort only when contamination exceeds an explicit threshold.

Tools: freeAny AI
Data engineering158words5 steps

Exclude frozen dates from market histories

Use a same-date reference bar to prevent closed-market snapshots from becoming new observations.

Tools: freeAny AI
Data engineering114words4 steps

Assign Table Values by Header Geometry

Map numeric cells to header positions so extra columns cannot silently shift field assignments.

Tools: freeAny AI
Data engineering134words5 steps

Apply the Same Price Floor to Every Served Query

Prevent raw-table loaders from bypassing the price contract used by authoritative views.

Tools: freeAny AI
Data engineering115words4 steps

Resolve email fields with a cost-aware source ladder

Use held data first, join existing records, and call a paid source only for a specific gap.

Tools: freeAny AI
Data engineering175words6 steps

Keep provisional values inside editable fields

Prevent a convenient starting value from becoming an uncorrectable factual claim.

Tools: freeAny AI
Data engineering142words5 steps

Discover Listing Pages and Validate Join Keys

Use list view to expose detail links, then verify URL identifiers against displayed listing numbers before joining records.

Tools: freeAny AI
Data engineering176words6 steps

Check Whether a Repair Survives Synchronization

Rerun the upstream writer, detect restored errors, and correct ownership instead of entering a write loop.

Tools: freeAny AI
Data engineering122words5 steps

Add incremental extraction per source

Classify each source before choosing a cursor, merge, or full replacement strategy.

Tools: freeAny AI
Data engineering251words7 steps

Enrich the scope actually present in a widened pool

Use records' own scope keys after fallback expands the result set.

Tools: freeAny AI
Data engineering99words5 steps

Cross-check ingestion identities in CI

Detect configuration drift by comparing registry entries with code, workflows, consumers, and landed data.

Tools: freeAny AI
Data engineering173words6 steps

Use direct counts for dataset-size claims

Replace stale planner estimates with a live count of the dataset being described.

Tools: freeAny AI
Data engineering110words4 steps

Correct Calendar-Based Year-Over-Year Specs

Replace row-offset comparisons with the specified calendar tolerance rule before implementation.

Tools: freeAny AI
Data engineering123words5 steps

Confirm a Data Source’s Full Scope

Enumerate a bounded source family and verify ingestion coverage against its authoritative catalog.

Tools: freeAny AI
Data engineering164words6 steps

Escape CSV cells at one output boundary

Route every exported cell through the tested escaper and verify formula-shaped input remains text.

Tools: freeAny AI
Data engineering131words5 steps

Classify Off-Market Listing Transitions

Resolve listing departures without mistaking pending closures or old sales for current outcomes.

Tools: freeAny AI
Data engineering203words6 steps

Show Price Basis in Comparable Bar Labels

Build sorted price bars with visible estimate and listing qualifiers, then attach them directly to the answer.

Tools: freeAny AI
Data engineering200words7 steps

Preserve feed identity through calculations

Carry source-feed labels through filtered history and derived values instead of hardcoding provenance.

Tools: freeAny AI
Data engineering120words5 steps

Cache and Reconcile Citation Detail

Replace repeated full citation scans with one reusable extract, guarded by exact count reconciliation.

Tools: freeAny AI
Data engineering140words5 steps

Build and Verify a Corridor Profile

Capture context, assemble evidence, preview generated sections, validate, and verify the saved profile.

Tools: freeAny AI
Data engineering182words6 steps

Build an Aligned Two-Tier Monthly Dataset

Preserve unmatched tiers while aligning monthly series and merging them at a stable grain.

Tools: freeAny AI
Data engineering160words6 steps

Keep data registries wired to consumers

Record source coverage, authoritative roots, obligations, and their live consumers.

Tools: freeAny AI
Data engineering152words6 steps

Build a Producer from Recorded Source Bodies

Replay exact fetched bodies offline and bind each extracted field to its actual retrieval.

Tools: freeAny AI
Data engineering158words7 steps

Bind chart and headline to one source total

Generate a headline and labeled donut from one detail table, then align preview and live output.

Tools: freeAny AI
Data engineering185words5 steps

Backfill Missing Listing Dates Safely

Recover missing listing dates in resumable batches without generating listing activity.

Tools: freeAny AI
Data engineering275words7 steps

Audit Profiles With Documented Exceptions

Separate complete profiles, justified nulls, and missing coverage without inventing measurements.

Tools: freeAny AI
Data engineering198words6 steps

Attribute Usage to Account Owners

Make account limits count the same owner's actions across web and tool interfaces.

Tools: freeAny AI
Data engineering171words6 steps

Apply a Prepared Results-Access Migration

Run an existing access-policy migration and verify authenticated users can see newly generated results.

Tools: freeAny AI
Data engineering107words4 steps

Calculate ratios only on matching dates

Prevent a lagging series from silently producing a ratio between different observation dates.

Tools: freeAny AI
Data engineering116words4 steps

Add a Domain to a Data Brief Factory

Extend a shared brief pipeline with a source adapter and deterministic analysis, preserving fact provenance.

Tools: freeAny AI
Data engineering212words6 steps

Verify the reply inbox end to end

Prove replies arrive in a readable inbox before depending on reply-based support or stop requests.

Tools: freeAny AI
Email production116words5 steps

Validate email rendering before delivery

Test the real builder and renderer, then inspect both layout and factual output.

Tools: freeAny AI
Email production129words5 steps

Reject misleading chart input mappings

Prevent negative bars and categorical labels from acquiring false visual or date meanings.

Tools: freeAny AI
Email production113words4 steps

Validate cached prose against the receiving artifact

Reject cached commentary containing numbers unsupported by its smaller destination.

Tools: freeAny AI
Email production80words5 steps

Trace every rendered number to a source fact

Bind typed facts into fixed output slots and lint the final artifact.

Tools: freeAny AI
Email production149words6 steps

Trace Field Preservation Through Document Builders

Follow complete input facts into document blocks and identify the first missing value or unverified boundary.

Tools: freeAny AI
Email production161words5 steps

Design email recipes around one pipeline

Let recipes choose content while a shared seam assigns layout and one renderer produces output.

Tools: freeAny AI
Email production121words5 steps

Render email and social charts from one specification

Compute geometry in code and feed identical chart specifications into each surface's image renderer.

Tools: freeAny AI
Email production117words5 steps

Share expanded brand tokens across email and social

Resolve typography, colors and surfaces from one branding record for both output types.

Tools: freeAny AI
Email production124words4 steps

Seed required blocks before applying brand data

Mirror the production canvas when a recipe reads brand information from existing blocks.

Tools: freeAny AI
Email production117words5 steps

Save Before Sending Personalized Email

Send the persisted design and substitute recipient fields separately for each recipient.

Tools: freeAny AI
Email production114words5 steps

Route Email Builds Through a Total Fallback

Make every email build use a registered grid, including unmatched requests and invalid builder results.

Tools: freeAny AI
Email production180words6 steps

Extend equivalent emails through the same builder

Change declared variant fields instead of copying a sibling email's construction logic.

Tools: freeAny AI
Email production119words6 steps

Render sentence banks without partial claims

Fill typed slots from resolved facts, drop incomplete sentences whole, and keep generated prose additive.

Tools: freeAny AI
Email production208words6 steps

Bake an email photo badge into the image

Composite the badge server-side and preserve the original photo if processing fails.

Tools: freeAny AI
Email production112words5 steps

Require a submitted form to unsubscribe

Make unsubscribe links display a confirmation form so automated link fetches do not remove recipients.

Tools: freeAny AI
Email production108words4 steps

Preserve supplied text outside AI rewrite slots

Protect authorized verbatim blocks from model edits, truncation, and schema caps.

Tools: freeAny AI
Email production105words5 steps

Retain debt for failed email delivery

Enforce the promised destination, confirm publication before sending, and preserve failed delivery for retry.

Tools: freeAny AI
Email production167words5 steps

Preserve Distinct Image Slots

Tag generated images by role so refreshing one preserves other images and manual uploads.

Tools: freeAny AI
Email production105words5 steps

Carry reader requirements into a new email product

Turn formatting and reader expectations into explicit acceptance checks for the new output.

Tools: freeAny AI
Email production122words5 steps

Fit chart labels with real font widths

Replace character-count clipping with measurements from runtime-accessible fonts.

Tools: freeAny AI
Email production83words5 steps

Parameterize repeated acceptance harnesses

Share verification logic while preserving each output's fixture and assertions.

Tools: freeAny AI
Email production95words5 steps

Render previews at the real output width

Prevent iframe width and scrollbars from fabricating wrapping defects.

Tools: freeAny AI
Email production83words5 steps

Apply Branding Without Changing Facts

Use one brand mapping for live previews while leaving generated factual content untouched.

Tools: freeAny AI
Email production106words5 steps

Make acceptance renders safe and correctly located

Ensure a gate's prescribed render writes its actual target without enabling paid lookups.

Tools: freeAny AI
Email production90words5 steps

Mirror the hero and deduplicate its gallery

Keep the leading image durable without repeating it in supplementary photos.

Tools: freeAny AI
Email production132words6 steps

Key cached assets by their rendered bytes

Make visual changes produce new asset URLs even when the input specification stays unchanged.

Tools: freeAny AI
Email production90words5 steps

Gate personalized recurring email sends

Build per-recipient issues in bounded batches, preview them, and require review before a new recurring product sends.

Tools: freeAny AI
Email production185words7 steps

Gate a Data Brief Before Delivery

Validate inputs, render and score the complete brief, then refuse delivery whenever a required gate fails.

Tools: freeAny AI
Email production213words6 steps

Resolve Brands Through Validated Fixtures

Prefer trusted stored brands, retain live enrichment as a fallback, and report the origin of each selection.

Tools: freeAny AI
Email production236words6 steps

Fill a Fixed Email Skeleton From Sources

Reuse sourced filling while leaving unavailable figures open and recovering safely from sourcing failures.

Tools: freeAny AI
Email production167words5 steps

Extend a Typed Document Block

Register a new block across typing, validation, defaults, rendering, editing, and branding.

Tools: freeAny AI
Email production106words5 steps

Precompute validated email commentary

Cache checked prose for bounded subjects and prove the email builder consumes it.

Tools: freeAny AI
Email production139words5 steps

Build an email grid with consistent typography

Use one layout grammar and tokens that keep font size, weight and line height together.

Tools: freeAny AI
Email production130words5 steps

Check email colors, buttons and links

Keep meaning visible without color or images and verify every action reaches its promised destination.

Tools: freeAny AI
Email production141words5 steps

Record which builder produced each document

Validate recipe keys and carry truthful provenance across document versions.

Tools: freeAny AI
Email production137words5 steps

Verify branding through the real profile path

Render with an account profile to catch fields lost between account and project.

Tools: freeAny AI
Email production131words5 steps

Suppress restricted facts before narrative generation

Remove facts the recipe must not reveal from both the grid and model input.

Tools: freeAny AI
Email production124words4 steps

Fit footnotes by choosing complete alternatives

Preserve whole factual statements instead of cutting a caption through a number.

Tools: freeAny AI
Email production115words5 steps

Clear narrative slots before authoring

Prevent prefilled template text from bypassing a narrator that skips occupied blocks.

Tools: freeAny AI
Email production121words4 steps

Clamp scheduled sends to a local-time window

Enforce local sending hours in scheduling code and the worker, preserving valid requests and explaining adjustments.

Tools: freeAny AI
Email production177words5 steps

Choose charts that explain the actual subject

Use charts for numerical arguments; remove irrelevant charts, decorative two-value bars, and empty chart frames.

Tools: freeAny AI
Email production126words6 steps

Refresh chart assets before trusting screenshots

Distinguish a cached preview from the image currently stored at a reused URL.

Tools: freeAny AI
Email production121words5 steps

Keep branding from overwriting authored content

Apply brand defaults after authoring and prove nonempty user content survives.

Tools: freeAny AI
Email production118words5 steps

Bound optional photo mirroring time

Keep an external image copy from leaving a build waiting indefinitely.

Tools: freeAny AI
Email production91words5 steps

Gate Email Intake on Authentication and Consent

Block unauthenticated submissions before download and restrict replies to the original sender.

Tools: freeAny AI
Email production184words7 steps

Wire stored secrets into the consuming runtime

Keep credentials out of tracked content and verify that the workflow actually receives its secret-store bindings.

Tools: freeAny AI
Operations122words4 steps

Connect obligation intake to automatic completion

Require a completion signal or explicit exception, then run the draining sweep.

Tools: freeAny AI
Operations109words5 steps

Verify private error reporting and alerts

Capture application errors across runtimes without exposing diagnostics or request bodies.

Tools: freeAny AI
Operations138words5 steps

Verify a Private Deployment

Transfer only reviewed files, verify their integrity, and isolate credentials from public serving paths.

Tools: freeAny AI
Operations125words5 steps

Verify a disabled service actually stops

Check live process command lines and persistent configuration after disabling a workload.

Tools: freeAny AI
Operations129words5 steps

Verify Which Code Reaches Production

Trace deployments and inspect the production revision before accepting claims about live behavior.

Tools: freeAny AI
Operations101words5 steps

Verify Orders Before Releasing Repaired EPUBs

Match paid orders to accepted diagnoses and issue private delivery with explicit expiry and renewal rules.

Tools: freeAny AI
Operations259words8 steps

Offer Repairs Only After a Clean Validation

Gate repair offers on measurable improvements, prevent duplicate active quotes, and recognize free samples.

Tools: freeAny AI
Operations186words8 steps

Close alerts with an auditable receipt

Assign alerts an owner, deadline, and closure evidence; keep pending items visible until disposition is recorded.

Tools: freeAny AI
Operations116words5 steps

Verify assumption-change signals in primary documents

Separate documented contradictions from allegations and indirect indicators.

Tools: freeAny AI
Operations139words4 steps

Test local security during component outages

Disconnect dependencies separately and verify recording, visible faults, and bounded aircraft recovery.

Tools: freeAny AI
Operations150words5 steps

Test authorization with real unprivileged users

Probe ownership boundaries rather than counting policies or trusting privileged requests.

Tools: freeAny AI
Operations119words5 steps

Stop Launches When a Shared Quota Is Exhausted

Detect quota exhaustion, record recovery state, and suspend every session until the reported reset.

Tools: freeAny AI
Operations176words6 steps

Stage rate limits before enforcement

Observe legitimate traffic before enabling rejection at the outer rate-limit layer.

Tools: freeAny AI
Operations135words5 steps

Audit a small app across infrastructure layers

Prioritize observed cost, exposure and recovery gaps using evidence from the actual application.

Tools: freeAny AI
Operations125words5 steps

Serialize publishers before loading authority

Prevent delayed publishers from replacing newer approved output or reviving revoked routes.

Tools: freeAny AI
Operations161words5 steps

Check Producer Heartbeats on an Independent Schedule

Detect a silent producer without waiting for its next push.

Tools: freeAny AI
Operations128words4 steps

Deploy Compatibility Before Renaming Configuration

Keep dependent requests working while configuration names change across code and deployment settings.

Tools: freeAny AI
Operations121words5 steps

Separate user traffic from deployment probes

Verify that logged activity represents users before using it to justify caching work.

Tools: freeAny AI
Operations94words5 steps

Dispatch publication without rerunning collection

Let approval publish existing output without triggering unrelated source checks or notices.

Tools: freeAny AI
Operations96words4 steps

Select one artifact across every result page

Keep artifact identity fields together when choosing a retained build from paginated results.

Tools: freeAny AI
Operations113words5 steps

Schedule Visible Control-Branch Tests

Run synthetic control tests weekly and surface failures through the normal operational reporting channel.

Tools: freeAny AI
Operations162words5 steps

Run Smoke Tests Linked to Verification Records

Run concurrent HTTP assertions, stamp passing records, and expose failures and manual checks without closing them.

Tools: freeAny AI
Operations257words8 steps

Review an SVG Package Before Delivery

Check visual fidelity and delivery records before releasing an SVG archive.

Tools: freeAny AI
Operations179words5 steps

Resume setup after secret synchronization fails

Retry external synchronization independently of local creation and declare readiness only after a live baseline.

Tools: freeAny AI
Operations117words5 steps

Resume batch work through persisted handoffs

Persist batch identifiers and input fingerprints before polling so later runs can collect safely.

Tools: freeAny AI
Operations157words6 steps

Test writes and sequences after restoration

A restored app must create, serve and remove new data without reusing previously issued IDs.

Tools: freeAny AI
Operations124words4 steps

Reserve API budget before making calls

Refuse spending when limits or accounting state cannot be trusted.

Tools: freeAny AI
Operations126words5 steps

Distribute Channel Assets From One Report

Turn a sourced report into audience-specific assets and direct every drop to its live canonical page.

Tools: freeAny AI
Operations136words5 steps

Recover Hosted Links After a Tunnel Restart

Recover the current hostname, verify hosted routes, and replace stale links and delivery references.

Tools: freeAny AI
Operations169words6 steps

Rebuild Dashboards From Current Inputs

Preserve current tracked topic dashboards, restore non-topic inputs, and rebuild the published site.

Tools: freeAny AI
Operations138words5 steps

Queue Work Behind an Active Runner

Use a bounded detached waiter that retries the locked launch when capacity becomes available.

Tools: freeAny AI
Operations146words6 steps

Gate every publication on current approval

Keep observations separate from approved output and enforce current approval and authorization on every publishing path.

Tools: freeAny AI
Operations156words5 steps

Publish and Verify Before Sending Notices

Build and deploy the change, verify its live page, then send notices and commit delivery state.

Tools: freeAny AI
Operations150words5 steps

Run Private Intake With Preview and Safe Retries

Separate preview from sending, serialize processing, and retry failed replies without creating duplicate deliveries.

Tools: freePythonAny AI
Operations196words8 steps

Package a Scoped SVG Batch

Build a reviewable SVG archive while preserving originals and excluding private intake data.

Tools: freePythonAny AI
Operations182words5 steps

Detect stale outputs beyond green schedules

Measure the age of produced data and separately prove scheduled execution occurs.

Tools: freeAny AI
Operations120words5 steps

Gate and order nightly data workflows

Require landed rows before publishing, then reject downstream sends with stale publication tokens.

Tools: freeAny AI
Operations183words6 steps

Monitor Hosted Pages Through a Service Commitment

Check every required service and promised page daily through the agreed hosting period.

Tools: freeAny AI
Operations189words6 steps

Migrate DNS Without Dropping Records

Inventory every DNS record before changing nameservers, then verify the imported set after migration.

Tools: freeAny AI
Operations138words5 steps

Measure actual traffic before adding a cache

Target the path users call instead of optimizing a merely existing endpoint.

Tools: freeAny AI
Operations83words5 steps

Record out-of-band actions without repeating them

Keep a ledger accurate when a person completes work outside the application.

Tools: freeAny AI
Operations136words5 steps

Promote tasks to verification only after code lands

Keep an empty specification from masquerading as completed implementation.

Tools: freeAny AI
Operations76words5 steps

Check the operations index before building a report

Avoid duplicating a monitor that already answers the question.

Tools: freeAny AI
Operations80words5 steps

Check the work behind a parked workflow

Inspect concrete deliverables as well as a workflow's gate and trigger state.

Tools: freeAny AI
Operations104words5 steps

Add access barriers beyond row policies

Restrict network access and remove unused database grants before a policy change can expose them.

Tools: freeAny AI
Operations123words5 steps

Guard Collectors Using Removable Archives

Stop capture when the intended archive volume is absent or storage safeguards fail.

Tools: freeAny AI
Operations165words6 steps

Wait for the Tested Commit to Be Deployed

Release smoke tests only when a ready production deployment matches the workflow's triggering commit.

Tools: freeAny AI
Operations171words6 steps

Gate rollback on critical smoke failures

Retry critical checks once before a guarded rollback; report soft failures without rolling back.

Tools: freeAny AI
Operations160words5 steps

Gate a One-Shot Scheduled Send

Limit a scheduled message to one request and reconcile uncertain delivery before claiming contact.

Tools: freeAny AI
Operations164words5 steps

Gate Catalog Promotion on a Clean Live Cycle

Keep a new data product provisional until its live pipeline, independent parity, and output checks pass.

Tools: freeAny AI
Operations148words6 steps

Gate supervised drone missions

Block launches and route changes unless authorization, fresh safety inputs, and recovery capacity pass.

Tools: freeAny AI
Operations171words5 steps

Attach Evidence Atomically When Closing Checks

Evaluate stored signals before closure, require evidence for manual checks, and clear proof when reopening.

Tools: freeAny AI
Operations205words5 steps

Prove a backup restores in isolation

Restore a named archive, exercise it, record measured results and destroy the test environment.

Tools: freeAny AI
Operations111words4 steps

Build a Shared Cron Failure Classifier

Turn log tails into consistent classes and guarded suggestions without model calls.

Tools: freePythonAny AI
Operations201words6 steps

Separate deploy rollback from data recovery

Write distinct recovery paths for application code, database changes and stored files.

Tools: freeAny AI
Operations127words5 steps

Deploy Public Files to a Shared Subpath

Copy selected public assets into an assigned subdirectory, verify them, then publish its navigation link.

Tools: freeGitAny AI
Operations205words6 steps

Measure rates when judging a cumulative counter

Use interval deltas to decide whether a fix reduced ongoing usage.

Tools: freeAny AI
Operations123words5 steps

Automatically close verified obligations

Give an obligation ledger a scheduled closer that distinguishes completion from superficial matches.

Tools: freeAny AI
Operations117words5 steps

Check account constraints before restarting

Separate the layer reporting an outage from quota or account restrictions that may keep causing it.

Tools: freeAny AI
Operations102words5 steps

Retry Transient Cron Failures Once

Recover temporary failures while preventing retry loops and preserving stale-data alerts.

Tools: freeAny AI
Operations154words6 steps

Execute a Bounded Data Repair

Establish authority, reject stale plans, change only the intended field, and verify outcomes separately.

Tools: freeAny AI
Operations202words7 steps

Bind changing claims to their source of truth

Replace maintained literals with derived values and block changes that leave authoritative records stale.

Tools: freeAny AI
Operations125words5 steps

Bench-test drone charging interlocks

Require safe seating and battery conditions, then inject faults with motors physically disabled.

Tools: freeAny AI
Operations149words5 steps

Archive Obsolete Code Without Breaking Dependents

Move retired material out of active discovery while preserving files still required by imports, jobs, or checks.

Tools: freeAny AI
Operations155words6 steps

Archive old obligations without losing their record

Reset an authorized backlog while retaining complete history and reversible dispositions.

Tools: freeAny AI
Operations118words5 steps

Research & knowledge

Knowledge management · Research workflows · Web research · Source discovery43 files

Recover an unfamiliar product from session records

Search prior transcripts before replacing an unknown name with a web-search guess.

Tools: freeAny AI
Knowledge management89words5 steps

Route an AI to relevant notes on demand

Search a small keyword map and read the matching pages instead of preloading the whole wiki.

Tools: freeAny AI
Knowledge management134words5 steps

Review corrections without freezing observations

Detect meaningful changes under existing identifiers while retaining the approved display until review.

Tools: freeAny AI
Knowledge management145words5 steps

Save reviews where the next session can find them

Make reviews durable artifacts instead of leaving them inside one conversation.

Tools: freeAny AI
Knowledge management78words5 steps

Keep immutable sources beside a maintained wiki

Compile linked notes while preserving raw evidence and explicit supersession.

Tools: freeAny AI
Knowledge management122words5 steps

Count document reachability without self-credit

Exclude generated inventories from evidence that documents are meaningfully discoverable.

Tools: freeAny AI
Knowledge management75words5 steps

Separate quoted and proxy options backtests

Use covered bid/ask quotes and label uncovered pricing proxies without duplicating spread costs.

Tools: freeAny AI
Research workflows156words4 steps

Validate API filters before trusting absence

Check that a filter narrows results before treating an empty result as evidence.

Tools: freeAny AI
Research workflows132words5 steps

Screen Concentrated Citation Bursts

Combine short citation spans with country concentration to catch artifacts missed by entity diversity.

Tools: freeAny AI
Research workflows169words6 steps

Score candidate sources by outcomes

Label candidates before scoring, compare settled outcomes consistently, and base source status changes on recorded evidence.

Tools: freeAny AI
Research workflows133words5 steps

Score Backtests by Market Regime

Expose regime-dependent failures that aggregate backtest results can conceal.

Tools: freeAny AI
Research workflows154words5 steps

Keep the countercase in the linked artifact

Preserve each candidate's adversarial reasoning where the decision record actually points.

Tools: freeAny AI
Research workflows95words4 steps

Make annualization frequency explicit

Require periods per year and inspect existing library behavior before annualizing a statistic.

Tools: freeAny AI
Research workflows104words4 steps

Report walk-forward efficiency with sample size

Show efficiency alongside out-of-sample trade count without inventing an unsupported acceptance threshold.

Tools: freeAny AI
Research workflows87words4 steps

Repair feature warmup in walk-forward tests

Precompute causal features when short test slices prevent out-of-sample trades.

Tools: freeAny AI
Research workflows128words4 steps

Measure Repair Coverage in a Lab and Pilot

Test controlled failures first, then measure supported repairs on authorized real incidents.

Tools: freeAny AI
Research workflows196words7 steps

Count parameter trials that produced no trades

Preserve every tested parameter cell while keeping undefined results out of dispersion calculations.

Tools: freeAny AI
Research workflows114words5 steps

Produce a Novel, Actionable Research Digest

Compare ideas against existing records and retain only findings with a concrete consequence.

Tools: freeAny AI
Research workflows233words6 steps

Package research for an AI follow-up

Export the accepted brief, traceable evidence, unresolved questions, and an editable follow-up as a portable snapshot.

Tools: freeAny AI
Research workflows187words7 steps

Isolate Exploratory Arms from Registered Comparisons

Keep a new experiment's data and state separate while preserving explicit comparator inputs.

Tools: freeAny AI
Research workflows114words6 steps

Build Point-in-Time Historical Prediction Grades

Generate internal calibration data only from series that can be reconstructed without look-ahead.

Tools: freeAny AI
Research workflows199words5 steps

Freeze a Universe with a Venue Liveness Screen

Remove duplicate numeraires and structurally favored pegged assets, then freeze venue-active candidates.

Tools: freeAny AI
Research workflows153words6 steps

Evaluate Added Features Against a Seasonal Baseline

Test whether extra domains improve a forecast using an untouched chronological holdout.

Tools: freeAny AI
Research workflows174words6 steps

Choose a graph by content and revision coverage

Distinguish code indexes, documentary graphs, and diagrams before querying them.

Tools: freeAny AI
Research workflows100words6 steps

Define a Forecast That Can Be Scored

Freeze the future outcome, input cutoff, release policy, and grading rules before forecasting.

Tools: freeAny AI
Research workflows165words6 steps

Keep a journal of process and outcome

Record assumptions before a decision and review its reasoning separately from what happened.

Tools: freeAny AI
Research workflows133words5 steps

Compare Cross-Domain History with Matched Periods

Compare each series against its own equivalent period while preserving geography and coverage.

Tools: freeAny AI
Research workflows165words6 steps

Classify Tier Divergence Deterministically

Combine spread and lower-tier changes while preventing upper-tier growth from overriding weakness.

Tools: freeAny AI
Research workflows166words5 steps

Author an Evaluable Scenario Task

Build a product-research task with separate scenario, persona sampling, verification, and reporting inputs.

Tools: freeAny AI
Research workflows210words7 steps

Bound Simulation Claims by Feed and Venue Depth

Expose historical-feed limits and simulated fills that exceed the activity represented by that feed.

Tools: freeAny AI
Research workflows157words6 steps

Audit a backtest before reporting results

Check trades, costs, bias, validation, and run provenance before treating results as evidence.

Tools: freeAny AI
Research workflows222words8 steps

Add a project-fit column to research findings

Show which measured problem each candidate addresses, or explicitly record no fit.

Tools: freeAny AI
Research workflows82words5 steps

Detect search failures behind empty results

Record answering engines and response validity before treating no results as no evidence.

Tools: freeAny AI
Web research126words5 steps

Run private metasearch with a JSON interface

Configure a local search service, keep it private and prove real JSON responses before wiring agents.

Tools: freeAny AI
Web research129words5 steps

Choose HTML extraction or a JSON API

Inspect the real response and use a typed API when it supplies the required records.

Tools: freeAny AI
Web research136words5 steps

Space Crawls to Avoid False Outage Reports

Pace requests to each host before treating burst-induced timeouts as source downtime.

Tools: freeAny AI
Web research170words5 steps

Diagnose an empty scrape before giving up

Inspect status, content and the fetch path before labeling the whole site inaccessible.

Tools: freeAny AI
Web research153words5 steps

Keep child processes from eating a shell loop

Give each crawler its own stdin so the parent loop processes every input record.

Tools: freeAny AI
Web research136words5 steps

Compare search tools by citation fidelity

Choose grounded retrieval by inspectable claim citations rather than apparent answer quality.

Tools: freeAny AI
Web research139words6 steps

Choose polling cadence from source changes

Match refresh frequency and cache expiry to verified publication behavior.

Tools: freeAny AI
Source discovery131words5 steps

Qualify a Data Source Before Adopting It

Bound access probes and establish meaning, coverage, timing, operating cost, and permitted use.

Tools: freeAny AI
Source discovery188words7 steps

Poll Listings With Control Gates

Detect new candidates while keeping failed polls from corrupting state or producing misleading silence.

Tools: freeAny AI
Source discovery178words7 steps

Evaluate an archive source before outreach

Assess public activity, ownership and contactability without collecting the archive itself.

Tools: freeAny AI
Source discovery144words5 steps

Practical reference

Website building30 files

Durable TypeScript workflows with the Workflow SDK

Builds durable TypeScript workflows with automatic retries, state persistence, resource-free sleeping, and human-in-the-loop approval hooks.

Tools: freeNodevercel account
Website building341words12 steps

n8n pipeline into Supabase storage and tables

Builds an n8n workflow to generate social media content and images using OpenAI, saving assets to Supabase.

Tools: freen8n accountsupabase account
Website building419words18 steps

Native CSS replacements for common JavaScript UI features

Implement scroll offsets, snapping, form validation styles, focus rings, and auto-growing textareas using pure CSS instead of JavaScript.

Tools: freeAny AI
Website building367words6 steps

AssemblyAI dictation and voice agents

Integrates speech-to-text, dictation, and voice agents into applications using AssemblyAI APIs.

Tools: paidGitNode
Website building230words9 steps

CSS-only details and summary animation

Animate HTML details disclosure elements and staggered child content using modern CSS transitions and starting-style rules.

Tools: freeAny AI
Website building371words11 steps

CSS text-box-trim and text-box-edge

Trims internal font whitespace to achieve exact, consistent spacing and alignment in web layouts.

Tools: freeAny AI
Website building237words5 steps

Local-first Next.js and Supabase with Claude Code

Set up a local-first Next.js and Supabase development workflow using Claude Code and Agent Skills.

Tools: freeNodeDocker
Website building424words13 steps

Next.js VPS deployment with Claude Code

Deploys a Next.js application to an Ubuntu VPS using Docker, GitHub Actions CI/CD, Caddy, and Cloudflare.

Tools: paidSSH accessDocker
Website building679words26 steps

Supabase pgvector setup

Enable the pgvector extension in Supabase and create a table with an embedding column for semantic and hybrid search.

Tools: freesupabase accountAny AI
Website building134words3 steps

CSS sibling-index and sibling-count functions

Animate and stack sibling elements dynamically without manual nth-child rules or JavaScript.

Tools: freeAny AI
Website building142words3 steps

Supabase Row Level Security and roles

Implement multi-tenancy, role-based access control, Row Level Security policies, and custom JWT claims in Supabase and PostgreSQL.

Tools: freesupabase accountAny AI
Website building517words14 steps

Supabase row-level security for anonymous users

Restrict each task to its anonymous owner using database RLS, shared sessions, and per-user foreign keys.

Tools: freeNodeDocker
Website building461words10 steps

Arcjet API endpoint protection for AI apps

Limits AI API usage and blocks bots, attacks, and sensitive data before costly model calls.

Tools: paidNodeOpenAI API key
Website building421words12 steps

Antigravity AI scroll-animation website workflow

Generate product visuals, convert video frames, and build a polished mouse-scroll website with AI.

Tools: paidAny AI
Website building416words10 steps

Stripe subscription sync pattern

Build reliable Stripe subscriptions with webhook-triggered pulls, local caching, checkout metadata, and server-side verification.

Tools: freeNodestripe account
Website building542words11 steps

Supabase Auth authentication and row-level security

Authenticate users with Supabase and authorize database rows using JWT-based row-level security policies.

Tools: freeAny AI
Website building405words12 steps

OWASP Top 10 2025 review checklist

Walk an application through the ten 2025 OWASP risk categories with a check and a mitigation for each.

Tools: freeAny AI
Website building371words10 steps

Supabase Edge Functions CLI guide

Create, test, deploy, invoke, package, and secure Supabase Edge Functions using TypeScript and Deno.

Tools: freesupabase-cliDocker
Website building445words16 steps

Vercel Fluid compute for Functions

Turn on Fluid compute for Vercel Functions so one instance serves concurrent requests, then check billing and shared state.

Tools: freevercel accountAny AI
Website building371words7 steps

React Scan render profiler

Finds excessive React renders and the components causing performance slowdowns.

Tools: freeNodeAny AI
Website building331words11 steps

UI design heuristics for AI development

Improve AI-generated user interfaces by eliminating visual clutter, fixing visual hierarchy, and avoiding default styling patterns.

Tools: freeNodeAny AI
Website building208words0 steps

Web accessibility auditing and HTML remediation

Detect and resolve web accessibility issues including landmarks, headings, alt text, contrast ratios, and keyboard navigation.

Tools: freeAny AI
Website building449words12 steps

Landing page hero section design framework

Structure landing page copy and select matching key visuals to improve visitor comprehension and conversion.

Tools: freeAny AI
Website building428words7 steps

Aura AI UI design editor

Generate and customize web UI layouts from reference screenshots without manual styling or repetitive code changes.

Tools: freeAny AI
Website building425words11 steps

Claude Design web design workflow

Build and iterate on web landing pages and custom design systems using Claude Design and visual references.

Tools: freeAny AI
Website building403words15 steps

Modern CSS layout patterns and features

Build responsive scrollers, adaptive auto-grids, container query components, and scroll animations using pure CSS.

Tools: freeAny AI
Website building388words10 steps

CSS spacing and typography layout

Configures vertical rhythm, flow spacing, container padding, grid gaps, and text trimming across web components.

Tools: freeAny AI
Website building377words6 steps

Taking the generic look out of AI web design

Eliminate generic AI aesthetics from generated landing pages using structured design systems, reference screenshots, and custom assets.

Tools: freeNodeAny AI
Website building397words11 steps

Chrome DevTools Performance panel

Diagnose web performance bottlenecks, replicate real-world user conditions, and test fixes directly within Chrome DevTools.

Tools: freeAny AI
Website building365words14 steps

Build a custom site with Claude Code and ship it

Install two front-end design skills, prompt for a site with animations, polish it in batches, then zip it and upload it to a host.

Tools: paidNodeClaude Code
Website building568words22 steps

Monitor a source and explain what changed

Build a baseline, detect meaningful changes, and produce an updated dashboard with an alert decision.

Tools: free
Task groups4steps

Give an AI a usable project memory

Compile supplied documents into a source-linked wiki and retrieve the right evidence for a question.

Tools: free
Task groups4steps

Bring a source into an AI data product

Implement ingestion whose records survive bad refreshes and are actually queryable by the intended AI consumer.

Tools: free
Task groups5steps

Produce reports whose claims trace to data

Generate a report or chart-backed answer using computed facts and auditable relationships.

Tools: free
Task groups5steps

Build consistent branded emails

Turn approved content into a branded email preview with checked layout, links and factual content.

Tools: free
Task groups5steps

Make an AI coding change with working checks

Map a requested change, implement its controls and prove the actual user path handles the target failure.

Tools: free
Task groups5steps

Prove an application can recover

Produce a recovery runbook backed by an isolated restoration and successful application writes.

Tools: free
Task groups3steps

Repair a listing value without losing the correction

Prepare and execute a scoped repair, verify the rendered result, and prove it survives upstream synchronization.

Tools: free
Task groups5steps

Publish exactly the content that was approved

Build a publication workflow in which corrections remain reviewable and delayed publishers cannot restore obsolete or revoked content.

Tools: free
Task groups4steps

Audit historical experiments without hiding failed trials

Produce an experiment audit that preserves every trial and tests whether historical signals used only available information.

Tools: free
Task groups4steps

Retrieve evidence and diagnose empty results

Choose a suitable retrieval route and distinguish missing evidence from broken extraction or ignored filters.

Tools: free
Task groups4steps

Know whether an automated job really worked

Instrument an existing job so missed runs, invalid output, resource problems and unresolved alerts are visible separately.

Tools: free
Task groups4steps