claude-code-toolkit
Verified coding workflows with routing, gates, and hooks
Instructs the agent to act as an orchestrator that routes requests to specialized agents and skills, loads context on demand, delegates deterministic work to scripts, and enforces a dense writing standard and untrusted-content boundary.
All Skills in This Repository (136)
Pure Emerald Level Indicatorsplanning-with-files
Persist plans and notes to markdown files for multi-phase tasks.
pipeline-scaffolder
Generate pipeline components from Pipeline Spec JSON into skills, agents, references, scripts, and routing.
comment-quality
Detect temporal wording in code comments and generate timeless rewrites with a structured report.
plan-manager
Manage plan lifecycles via CLI with validation and staleness checks.
post-outliner
Analyze topic briefs to create detailed blog outlines with frontmatter planning.
perses-datasource-manage
Manage Perses datasources across global, project, and dashboard scopes.
professional-communication
Extract propositions from technical updates into executive-ready summaries with status indicators.
typescript-check
Run tsc --noEmit and generate file-grouped TypeScript error reports.
perses-dac-pipeline
Automate Perses Dashboard-as-Code pipelines from module initialization to CI/CD deployment.
vitest-runner
Execute Vitest test suites and report file-level results with durations and failure traces.
perses-dashboard-review
Fetch and analyze Perses dashboards to generate structured quality reports.
pr-mining-coordinator
Coordinate GitHub PR mining jobs to extract coding rules and track progress.
Frequently Asked Questions
FAQPage SchemaHow to install claude-code-toolkit?โผ
Run `npx skills add notque/claude-code-toolkit --all -g -y` in your terminal to install the full toolkit globally.
What does claude-code-toolkit do?โผ
It routes your plain-English requests to the right specialist agent and workflow skill, then enforces verification gates so work ships with test evidence instead of unverified claims.
How does it stop agents from skipping tests?โผ
Hooks fire automatically on lifecycle events and block completion until exit codes and test output prove the work, so 'looks correct' never passes as done.
Does it work with Codex and other harnesses?โผ
Yes. The installer mirrors agents, skills, and supported hooks into Codex, Factory, and Reasonix, with the same /do router entry point in each.
Can non-engineers use claude-code-toolkit?โผ
Yes. Beyond coding, it includes research, fact-checking, writing, and HTML artifact skills, all triggered by plain-English requests with no setup knowledge required.
Related Repositories in Software Engineering
View All in Software Engineeringโopenclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core