Core development loop: code review, fixes, commits, linting hooks, and 24 language-specific review agents. Skills available for both Claude Code and Codex CLI; agents, hooks, and commands are Claude Code-only.
| Skill |
Invocable |
What It Does |
ccgram-messaging |
yes |
Inter-agent messaging via ccgram swarm |
committing-code |
yes |
Smart git commits with logical grouping |
reviewing-code |
yes |
Multi-agent review (security, quality, idioms) |
fixing-code |
yes |
Parallel agents fix all issues, zero tolerance |
documenting-code |
yes |
Update docs based on recent changes |
improving-tests |
yes |
Refactor tests: combine to tabular, fill gaps |
refactoring-code |
auto |
Multi-file batch changes via MorphLLM |
searching-code |
auto |
Intelligent codebase search via WarpGrep |
coding |
auto |
Process discipline: surface assumptions, define verifiable goals before coding |
docs-keeper — documentation updates
- 6 Go review agents:
go-qa, go-tests, go-idioms, go-impl, go-simplify, go-docs
- 6 Python review agents:
py-qa, py-tests, py-idioms, py-impl, py-simplify, py-docs
- 6 TypeScript review agents:
ts-qa, ts-tests, ts-idioms, ts-impl, ts-simplify, ts-docs
- 6 Web review agents:
web-qa, web-tests, web-idioms, web-impl, web-simplify, web-docs
| Hook |
Event |
What It Does |
skill-enforcer.sh |
UserPromptSubmit |
Pattern-matches prompt and suggests skills |
file-protector.sh |
PreToolUse |
Blocks edits to settings.json, secrets |
git-guardrails.sh |
PreToolUse |
Blocks destructive git commands |
smart-lint.sh |
PostToolUse |
Auto-runs linter after file edits |
test-runner.sh |
PostToolUse |
Auto-runs tests after implementation changes |
session-start.sh |
SessionStart |
Shows git branch, last commit, file context |
notify.sh |
Notification |
Desktop notifications for long operations |
performance-monitor.sh |
PostCompact |
Tracks context compaction metrics |
smart-lint.sh sources .claude-hooks-config.sh from the project root if present, allowing per-project overrides for linter settings, language toggles, and quality thresholds.
| Server |
Used For |
| Context7 |
Doc lookup in review agents |
| Perplexity |
QA and simplify agents check best practices |
| MorphLLM |
searching-code, refactoring-code |