claudex

Double-model verification gate before declaring a build/research/refactor "done". Runs a fresh Anthropic (Claude) review AND an independent OpenAI/Azure critic against the stated goal + criteria, then reconciles. Use before announcing completion of medium/major builds, refactors, major fixes, research deliverables, frontend work, or trade-decision rigor — and on demand via /verify. Skip for small Q&A and trivial edits.

Record and retrieve Architecture Decision Records — log a decision's context, options, rationale, and tradeoffs, then list/search them. Use when making a non-trivial design or architecture decision, choosing between tools/approaches, or asked 'why did we do it this way?'

Queue an external action (Telegram message, shell command, GitHub comment, PR merge, file write) for the user's explicit approval before it runs. Use when you want to take an outward-facing or irreversible action that needs sign-off first.

Run a structured multi-position debate over an analytical question, spawning sub-agents to argue each option and counter confirmation bias. Use when comparing options or tradeoffs — 'which is better', 'should we', 'what's the risk', 'evaluate', 'compare options'.

Ingest documents (URLs, local files, PDFs) into a searchable local knowledge base and query stored knowledge. Use when ingesting or storing documents, searching previously stored knowledge, or building research context.

Collect macro market and economic-calendar data (yfinance + FairEconomy API) into a SQLite DB and generate a formatted macro briefing for Telegram. Use when producing a macro/markets briefing or collecting macro market data.

Watch any URL for meaningful changes — price moves, new releases (semver), hiring signals, content diffs, keyword appear/disappear — and alert. Use when tracking competitors, releases, or job postings, or detecting content changes on a web page.

Classify a task and pick the right Claude model tier (haiku/sonnet/opus) before significant work. Use every time before spawning a sub-agent or starting substantial work, to right-size the model to the task.

Configure Nginx or Caddy as a reverse proxy with SSL/TLS. Use when setting up reverse proxies, Let's Encrypt certificates, routing domains to backend services, load balancing, or serving static files.

Execute multi-step pipelines defined in JSON — sequential and parallel stages, dependency chains, conditionals, retries with backoff, per-step timeouts, variable passing, and run history. Use when running or orchestrating a multi-step JSON-defined pipeline.

Monitor GitHub repo health — CI failures on main, PRs awaiting review over 24h, PRs with failing checks, stale PRs — and alert. Use when checking repo/PR/CI health, investigating why an alert fired, or adding a repo to watch.

Create and maintain one-page project briefs (goal, success criteria, constraints, stack, status, next actions). Use when starting a new project, updating a project's status, or asked 'what's the status of X?'

Decompose a project into tasks with dependencies, critical path, and tracking before execution. Use at the start of every significant project before spawning sub-agents or writing code; also for project status, task failures, or changes affecting other tasks.

Collect research findings (source, URL, content, confidence, tags) across sources and compile them into a structured report. Use when doing multi-source research and you want findings saved as a structured, citable report.

Subscribe to RSS/Atom feeds with keyword filters and run digests delivered to Telegram. Use when subscribing to feeds, checking which feeds run, running a digest manually, or adjusting feed keywords.

Track equity sector rotation — collect sector ETF data and compute relative-strength heatmaps, sector leaders, and market cycle phase. Use when analyzing sector rotation, relative strength across sectors, or where we are in the market cycle.

Score the credibility of a web source by domain tier, content signals, and claim support; handles single or batch URLs. Use when judging whether a source is trustworthy, comparing sources, or validating research findings.

Create and manage systemd services, timers, and socket units. Use when creating background services or daemons, setting up scheduled timers, debugging failed units, viewing unit logs, or managing service dependencies.

Persistent JSON-backed job queue for scheduling shell commands without raw cron — supports one-shot (run at a time) and recurring (cron schedule) tasks. Use when asked to schedule or run a job at/every some interval, list or cancel queued tasks, or check task status.

Scan project dependencies for known security vulnerabilities (npm-audit style) across your projects and alert. Use when checking projects for vulnerabilities, after adding a new npm project, or reviewing the last scan results.

Show the auto-curated user profile (memory/USER.md) — who the user is, preferences, working patterns, recent corrections, and open threads — with last-updated, section counts, and a token-cap/prune check. Use when asked "what do you know about me", "show my profile", "whoami", or to review/prune the profile.

Estimate and rank what is consuming the context window — CLAUDE.md, MEMORY.md, USER.md, the 162 auto-loaded skill descriptions (ranked heaviest-first), and tool/MCP schemas. Use when asked about context bloat, token weight, "what's eating my context", which skills are heaviest, or with --cost for real spend.

Meta-skill for detecting when a solved problem is worth packaging as a reusable skill, then scaffolding it. Runs automatically after completing any non-trivial task. Also use when explicitly asked to create, draft, or propose a new skill from something just built.

Semantic memory search across all agent memories and conversation history. Use BEFORE answering questions about prior work, decisions, dates, people, preferences, projects, or past conversations. Also use when asked "do you remember", "what did we discuss", "when did we", etc.

Get current weather and forecasts. Use when the user asks about weather, temperature, or forecasts for any location.

Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in (single or multi-account), or reading/injecting/running secrets via op.

Accessibility auditing: WCAG compliance checking, contrast ratios, ARIA labels, keyboard navigation, semantic HTML, screen reader compatibility.

Autonomous API testing and evaluation. Tests any REST API for correctness, security, performance, error handling, and standards compliance. Discovers endpoints, probes with valid/invalid/edge-case payloads, checks auth, response times, injection vulnerabilities, and generates severity-scored reports with actionable fixes. Use before any API "done" claim.

Systematic REST and GraphQL API endpoint testing: happy path, error cases, auth flows, rate limits, schema validation. Use when: testing backend APIs, validating endpoint responses, checking error ...

Manage Apple Notes via the `memo` CLI on macOS (create, view, edit, delete, search, move, and export notes). Use when a user asks OpenClaw to add a note, list notes, search notes, or manage note folders.

Manage Apple Reminders via remindctl CLI (list, add, edit, complete, delete). Supports lists, date filters, and JSON/plain output.

Generate architecture diagrams from codebase analysis: Mermaid, PlantUML, or ASCII.

Static code analysis like a senior developer — cyclomatic/cognitive complexity, dead code detection, dependency graph analysis (circular deps, god modules), pattern consistency, error handling audit, code duplication detection. Use during refactors or before shipping.

Automated backup strategies: database dumps, file snapshots, incremental backups, verification, disaster recovery. Use when: setting up backup schedules, creating database backups, restoring from b...

Create, search, and manage Bear notes via grizzly CLI.

Monitor blogs and RSS/Atom feeds for updates using the blogwatcher CLI.

BluOS CLI (blu) for discovery, playback, grouping, and volume.

Use when you need to send or manage iMessages via BlueBubbles (recommended iMessage integration). Calls go through the generic message tool with channel="bluebubbles".

Read and manage Google Calendar events: list upcoming, create events, check availability, send reminders.

Capture frames or clips from RTSP/ONVIF cameras.