Language-agnostic engineering rules, review philosophy, and safety guards
npx claudepluginhub kiringyj/claude-workbench --plugin coreApply Linus Mode code review to staged or specified changes
Sync claude-workbench rules and plugin settings to this project
Review and update project documentation to match current state
Review and update project TODO.md to reflect current implementation state
Centralized Claude Code plugin marketplace and shared rules for cross-project reuse.
This repo serves two purposes:
rules/) — synced to consumer projects as .claude/rules/*.mdplugins/) — functional extensions (skills, hooks, hookify rules, MCP servers)Rules are not inside plugins. Plugins provide functionality (hooks enforce checks, skills encode workflows); rules provide persistent context (conventions, philosophy, naming). Consumer projects sync both via a single skill.
| File | Scope |
|---|---|
rules/core.md | Language-agnostic engineering rules, review philosophy, naming conventions |
rules/rust.md | Rust naming, cargo workflow, pre-commit checks |
rules/python.md | Python naming, uv/ruff/ty workflow, logging conventions |
| Plugin | Description |
|---|---|
| core | Safety guards (hookify), doc review hook, code review skill, MCP servers |
| rust | Pre-commit guard (fmt/clippy/check), optimize skill |
| python | Pre-commit guard (ruff/ty), optimize skill, ty LSP server |
sync-workbench skill):claude plugin install core --marketplace workbench --source github:KiringYJ/claude-workbench
/sync-workbench rust
This generates .claude/rules/*.md and merges plugin settings into .claude/settings.json.
| Profile | Rules | Plugins |
|---|---|---|
base | core | core@workbench |
rust | core, rust | core@workbench, rust@workbench |
python | core, python | core@workbench, python@workbench |
full | core, rust, python | core@workbench, rust@workbench, python@workbench |
Re-run /sync-workbench at any time. It is idempotent — unchanged files are skipped, and existing settings (permissions, env, custom plugins) are preserved.
Flags:
--rules-only — only sync .claude/rules/*.md--settings-only — only sync .claude/settings.json--check — report whether anything is out of date without writingpython scripts/sync_workbench.py rust --source /path/to/claude-workbench --target /path/to/project
python scripts/sync_workbench.py --check --source /path/to/claude-workbench --target /path/to/project
workbench.json # profile definitions (rules + plugins per profile)
rules/
core.md # shared rules (synced to consumer projects)
python.md
rust.md
scripts/
sync_workbench.py # standalone sync script for CI / non-Claude usage
plugins/
core/
.claude-plugin/
plugin.json
hooks/
document_sync.py
enforce_rules.py
hooks.json
test_document_sync.py
test_enforce_rules.py
skills/
review/
SKILL.md
sync-workbench/
SKILL.md # sync rules + settings to consumer projects
update-document/
SKILL.md
update-todo/
SKILL.md
.mcp.json
hookify.block-bad-git-add.md
hookify.block-compound-cd.md
hookify.block-no-verify.md
hookify.ultrathink-mode.md
python/
.claude-plugin/
plugin.json
hooks/
guard-python-commit.sh
hooks.json
skills/
optimize/
SKILL.md
rust/
.claude-plugin/
plugin.json
hooks/
guard-rust-commit.sh
hooks.json
skills/
optimize/
SKILL.md
.claude/
skills/
new-plugin/
SKILL.md
plugin-test/
SKILL.md
release/
SKILL.md
MIT
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Matches all tools
Hooks run on every tool call, not just specific ones
Executes bash commands
Hook triggers when Bash tool is used
Search and retrieve documents from local markdown files.
Intelligent prompt optimization using skill-based architecture. Enriches vague prompts with research-based clarifying questions before Claude Code executes them
Streamline people operations — recruiting, onboarding, performance reviews, compensation analysis, and policy guidance. Maintain compliance and keep your team running smoothly.
Prospect, craft outreach, and build deal strategy faster. Prep for calls, manage your pipeline, and write personalized messaging that moves deals forward.
Qiushi Skill: methodology skills for AI agents guided by seeking truth from facts, with Claude Code, Cursor, OpenClaw, Codex, OpenCode, and Hermes guidance.