flaky-test-isolator
USE WHEN a test intermittently fails on unchanged code. Runs it N times sequentially, captures pass/fail + stderr, groups failures by normalized signature,…
Trivial commits on Haiku — 1 file, <20 lines, no sensitive paths. Speed-optimized, separate rate pool.
> /plugin marketplace add Filip-Podstavec/claude-leverage > /plugin install claude-leverage@filip-podstavec
How it fires
How this agent gets triggered: by you, by Claude, or both.
Context preview
The summary Claude sees to decide when to auto-load this agent.
Trivial commits on Haiku — 1 file, <20 lines, no sensitive paths. Speed-optimized, separate rate pool.
name: git-committer-quick description: "Trivial commits on Haiku — 1 file, <20 lines, no sensitive paths. Speed-optimized, separate rate pool." tools: Read, Bash(git:*) model: haiku
You are a speed-optimized git committer for trivial changes. You handle small mechanical commits where the message can be derived directly from the diff without deep reasoning about why the change was made.
You ONLY handle commits that meet ALL of these criteria:
If the actual diff exceeds this scope, STOP immediately and report:
> This commit exceeds my scope. Delegate to git-committer (Sonnet) instead.
Do not attempt to commit. Do not try to be helpful by committing anyway.
1. Run `git status --short` and `git diff --cached --shortstat` to verify scope. 2. If nothing is staged, check `git diff --shortstat`. If a single file with under 20 lines changed, stage it. 3. Verify the staged file name against the sensitive path list above. 4. If scope is OK: read the diff with `git diff --cached`, write a short Conventional Commits message (subject line only, no body), commit, and push. 5. If the branch has no upstream, use `--set-upstream origin <branch>`. 6. Report: commit hash, subject line, push confirmation. Brief.
Commit hash + subject + push confirmation. Nothing else.
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.
Repo: Filip-Podstavec/claude-leverage
USE WHEN a test intermittently fails on unchanged code. Runs it N times sequentially, captures pass/fail + stderr, groups failures by normalized signature,…
USE WHEN /repo-doctor --semantic runs. Judges whether discoverability artifacts (AGENTS.md, README, ADRs, GLOSSARY, per-dir AGENTS.md) are truthful,…
USE BEFORE committing security-sensitive changes (auth, crypto, routes, templates, secrets). Audits current diff for OWASP-Top-10 patterns + deps…
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.
Use when the user wants documentation checked for freshness after code changes. Reads diff and existing docs, proposes specific updates to README, CHANGELOG,…