Agent Claude Code
Run lint, build, and tests on demand and report pass/fail with the relevant errors. Use when validating that a change compiles and the suite is green.
2 9d ago A 36 tokens
original MIT
CodeReviewer
02Agent Claude Code
Review a diff and return a summary plus issues, making no edits. Use for reviewing a branch, PR, or working-tree change.
2 9d ago A 30 tokens
original MIT
DocWriter
03Agent Claude Code
Maintain documentation - keeps decisions.md, errors.md, and CLAUDE.md files accurate. Use when docs have drifted from the code.
2 9d ago A 31 tokens
original MIT
LogAnalyzer
04Agent Claude Code
Parse crash logs, build errors, and stack traces to surface the root cause. Use when handed a failure log to triage.
2 9d ago A 29 tokens
original MIT
ReadOnly
05Agent Claude Code
Safe read-only codebase exploration with zero edits. Use to locate and understand code without any risk of modification.
2 9d ago A 25 tokens
original MIT
Researcher
06Agent Claude Code
Fetch and synthesize web sources, returning findings rather than opinions. Use for external research on libraries, APIs, or prior art.
2 9d ago A 29 tokens
original MIT
batch
07Command Claude Code
Split a large task into parallelisable subtasks and delegate to subagents.
2 9d ago A 0 tokens
original MIT
checkpoint
08Command Claude Code
Write current session state to .claude/checkpoint.md. Run before ending a mid-task session or when context is running low.
2 9d ago A 0 tokens
original MIT
commit
09Command Claude Code
Stage all changes, generate a concise commit message, commit, and push.
2 9d ago A 0 tokens
original MIT
plan
10Command Claude Code
Scaffold a structured plan file for a task or feature.
2 9d ago A 0 tokens
original MIT
pr
11Command Claude Code
Create a pull request from current branch to main.
2 9d ago A 0 tokens
original MIT
review
12Command Claude Code
Invoke the CodeReviewer agent on the current diff. Ask permission before spawning.
2 9d ago A 0 tokens
original MIT
ship
13Command Claude Code
Full ship sequence: lint → build → test → commit → PR. Stops on first failure.
2 9d ago A 0 tokens
original MIT
verify
14Command Claude Code
Command "verify" from moneytosms/claude-start, covering /verify, what to add here, template, /verify and steps.
2 9d ago A 0 tokens
original MIT
SessionStart
15Hook Claude Code
Runs when a session starts, executing session-start.sh via bash. From moneytosms/claude-start.
2 9d ago A
tokens not measured
original MIT
Notification
16Hook Claude Code
Runs when the agent sends a notification for idle_prompt and permission_prompt tool calls, executing notify.sh via bash. From moneytosms/claude-start.
2 9d ago A
tokens not measured
copy · 83% MIT
PreToolUse
17Hook Claude Code
Runs before the agent uses a tool for Bash, Read and Bash tool calls, executing log-bash.sh, pre-deploy-guard.sh, git-guard.sh and privacy-guard.sh via bash (4 commands). From moneytosms/claude-start.
2 9d ago A
tokens not measured
copy · 89% MIT
PostToolUse
18Hook Claude Code
Runs after a tool call finishes for Edit, Write and NotebookEdit tool calls, executing auto-format.sh via bash. From moneytosms/claude-start.
2 9d ago A
tokens not measured
copy · 89% MIT
Stop
19Hook Claude Code
Runs when the agent finishes a response, executing keep-going.sh via bash. From moneytosms/claude-start.
2 9d ago A
tokens not measured
copy · 83% MIT
claude-start
20Settings file Claude Code
Agent settings declaring 5 hook events (SessionStart, Notification, PreToolUse, PostToolUse) and 15 allowed tools.
2 9d ago C
tokens not measured
original MIT
Instructions file
Instructions for moneytosms/claude-start, covering stack, canary, tooling, context rules and error protocol.
2 9d ago A 749 tokens
original MIT