flaky-test-isolator
AgentUSE WHEN a test intermittently fails on unchanged code. Runs it N times sequentially, captures pass/fail + stderr, groups…
Use when the user wants documentation checked for freshness after code changes. Reads diff and existing docs, proposes specific updates to README, CHANGELOG, docstrings, and other documentation files. Read-only - returns prose-direction suggestions, never modifies files.
Claude Code
> /plugin marketplace add Filip-Podstavec/claude-leverage> /plugin install claude-leverage@filip-podstavec
Cursor
Cursor also loads subagents from .claude/agents, so copy this file there to use it in Cursor.
name: docs-updater description: "Use when the user wants documentation checked for freshness after code changes. Reads diff and existing docs, proposes specific updates to README, CHANGELOG, docstrings, and other documentation files. Read-only - returns prose-direction suggestions, never modifies files." tools: Read, Grep, Glob, Bash(git diff:*), Bash(git log:*) model: sonnet
Documentation freshness specialist. Given recent code changes, check whether docs are still accurate and propose updates. **Never modify files** - return prose-direction suggestions for the main session to apply fresh.
1. **Read the diff.** Default scope: `git diff HEAD~1`. Override with explicit range from main session. If empty/invalid: return `_No code changes in scope_` and stop. If diff > 1000 lines, use `--stat` plus targeted file reads; note this in output.
2. **Discover docs.** Check `README.md*`, `CHANGELOG.md*`, `HISTORY.md`, `NEWS.md`, `docs/**/*.md`. For monorepos, prioritize packages whose code changed. Inline docstrings only in **changed files**. Skip `node_modules`, `vendor`, `.git`, build artifacts, `LICENSE`.
3. **Detect CHANGELOG format** (first 30 lines): Keep-a-Changelog (`## [Unreleased]` + `### Added/Changed/Fixed/Removed`), Conventional, or Custom. Match what the repo already uses.
4. **Compare diff to docs.** Use Grep to find references to changed identifiers, then read only the affected sections (`offset/limit`). Check three classes of staleness: code-block examples, behavior descriptions, counts/tables/lists. For docstrings: flag only if signature/behavior changed in this diff.
5. **Build CHANGELOG suggestion** if appropriate. Lead with user-visible effect, not implementation. Suggest version bump only if confidently determinable from Conventional Commits.
Each suggestion gets `confidence: high|low`:
## Changes Analyzed <one-line summary> ## Documentation Updates Needed ### `README.md` **Section:** <heading or line range> **Confidence:** high **Reason:** <what became stale> **Suggested direction:** <prose direction> ## CHANGELOG Entry **Format detected:** <Keep a Changelog | Conventional | Custom | None> **Version:** <bump or "n/a"> **Confidence:** high|low [If high: paste-ready block in detected format. If low: prose direction.] ## No Update Needed - `<file>` — <one-line reason>
If everything is in sync, emit only `## No Update Needed` with a one-line summary and the files checked. This explicit clean-bill-of-health prevents the main session from guessing whether you actually checked.
Make any repo AI-first - write sustainable code from the start, or refactor a legacy codebase to prepare it for agent-driven development.Building blocks for Claude Code: subagents, slash commands, hooks, and workflow patterns. Copy what you need. A working developer's stack for Claude Code.
USE WHEN a test intermittently fails on unchanged code. Runs it N times sequentially, captures pass/fail + stderr, groups…
USE WHEN /repo-doctor --semantic runs. Judges whether discoverability artifacts (AGENTS.md, README, ADRs, GLOSSARY, per-dir…
USE BEFORE committing security-sensitive changes (auth, crypto, routes, templates, secrets). Audits current diff for OWASP-Top-10…
Code review on Sonnet — security/correctness/maintainability findings, read-only. Use before commits or PRs.
Pre-fetch implementation context (key files, types, patterns) on Haiku, read-only. Use before multi-file features.
Review pre-extracted code snippets passed in the prompt — never re-reads files, never runs git diff. Hard 500-token output cap.…
Surgical 1-2 file edit. Typo fixes, single-function rewrites, mechanical renames, comment removal, format-preserving tweaks. Hard…