{"id":35734421,"url":"https://github.com/ait88/claude-workflow-toolkit","last_synced_at":"2026-01-13T20:40:26.552Z","repository":{"id":332025972,"uuid":"1128905110","full_name":"ait88/claude-workflow-toolkit","owner":"ait88","description":"Reusable workflow optimization toolkit for Claude Code agents.","archived":false,"fork":false,"pushed_at":"2026-01-12T05:20:55.000Z","size":81,"stargazers_count":0,"open_issues_count":3,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-01-12T10:30:41.373Z","etag":null,"topics":["agentic-workflow","claude-code","claude-skills","gh-cli","token-optimization"],"latest_commit_sha":null,"homepage":"","language":"Shell","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/ait88.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-06T10:15:03.000Z","updated_at":"2026-01-12T05:21:00.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/ait88/claude-workflow-toolkit","commit_stats":null,"previous_names":["ait88/claude-workflow-toolkit"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/ait88/claude-workflow-toolkit","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ait88%2Fclaude-workflow-toolkit","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ait88%2Fclaude-workflow-toolkit/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ait88%2Fclaude-workflow-toolkit/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ait88%2Fclaude-workflow-toolkit/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/ait88","download_url":"https://codeload.github.com/ait88/claude-workflow-toolkit/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/ait88%2Fclaude-workflow-toolkit/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28400040,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-13T14:36:09.778Z","status":"ssl_error","status_checked_at":"2026-01-13T14:35:19.697Z","response_time":56,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agentic-workflow","claude-code","claude-skills","gh-cli","token-optimization"],"created_at":"2026-01-06T12:13:39.130Z","updated_at":"2026-01-13T20:40:26.547Z","avatar_url":"https://github.com/ait88.png","language":"Shell","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Claude Workflow Toolkit\n\nA reusable template system for optimizing Claude Code and Codex agent workflows across any project.\n\n## What This Does\n\nThis toolkit provides:\n- **Workflow Skills**: Bash scripts that automate GitHub issue/PR workflows with 60-80% fewer API calls\n- **Review-First Workflow**: Skills that ensure review feedback is addressed before new work begins\n- **Security Checklists**: Tech-specific security validation guides\n- **Documentation Templates**: Quick-reference docs that help agents navigate your codebase efficiently\n- **Project Profiles**: Pre-configured settings for common tech stacks\n\n## How It Works\n\n1. **One-time application**: Run the toolkit against your project\n2. **Generates tailored files**: Skills, docs, and checklists customized to your project\n3. **Project owns the output**: Your repo maintains its own copy, independent of this toolkit\n4. **Optional updates**: Re-run later to pull in new optimizations\n\n## Quick Start\n\n### For Claude Code or Codex Agents\n\nIf you're a Claude Code or Codex agent asked to apply this toolkit:\n\n1. Read `SKILL.md` for detailed instructions\n2. Identify the target project's tech stack\n3. Select appropriate profile from `profiles/`\n4. Apply templates with project-specific values\n\n### For Humans\n\n1. Clone this repo alongside your project:\n   ```bash\n   git clone https://github.com/ait88/claude-workflow-toolkit.git ~/claude-workflow-toolkit\n   ```\n\n2. Ask your Claude Code or Codex agent:\n   ```\n   Apply the workflow toolkit from ~/claude-workflow-toolkit to this project.\n   Read SKILL.md in the toolkit for instructions.\n   ```\n\n3. Review and commit the generated files\n4. Optionally delete the toolkit clone (your project is now self-contained)\n\n## What Gets Generated\n\n```\nyour-project/\n├── .claude/\n│   ├── skills/              # Canonical skill scripts for both agents\n│   │   ├── check-reviews    # Find PRs needing response (run first!)\n│   │   ├── address-review   # Show review feedback for a PR\n│   │   ├── claim-issue      # Claim issue + create branch\n│   │   ├── check-workflow   # Validate workflow state\n│   │   ├── submit-pr        # Create PR + update labels\n│   │   └── README.md\n│   ├── commands/            # Skill documentation for discoverability\n│   │   ├── check-reviews.md\n│   │   ├── address-review.md\n│   │   ├── claim-issue.md\n│   │   ├── check-workflow.md\n│   │   └── submit-pr.md\n│   ├── SECURITY-CHECKLIST.md  # Tech-specific security guide\n│   └── settings.local.json    # Pre-configured Claude Code permissions\n├── .codex/\n│   └── skills -\u003e ../.claude/skills  # Mirror so Codex sees the same commands\n└── docs/\n    ├── QUICK-REFERENCE.md   # Fast navigation for agents\n    ├── FAQ-AGENTS.md        # Pre-answered questions\n    └── CODEBASE-MAP.md      # Annotated directory structure\n```\n\n### Dual-Agent Skills (Claude + Codex)\n\nKeep one canonical skills folder and mirror it for Codex so both agents run the same commands:\n\n```bash\nmkdir -p .codex\nrm -rf .codex/skills\nln -s ../.claude/skills .codex/skills  # adjust if you customize SKILLS_DIR\n```\n\n## Profiles\n\n| Profile | Best For |\n|---------|----------|\n| `default.yaml` | Any project |\n| `php-composer.yaml` | PHP with Composer |\n| `bash-cli.yaml` | Bash CLI tools |\n| `node-npm.yaml` | Node.js/TypeScript |\n| `python-poetry.yaml` | Python with Poetry |\n\n## API Efficiency\n\n### Traditional workflow (manual commands)\n- Check for reviews: 5+ API calls\n- Claim issue: 3-4 API calls\n- Check workflow: 4-5 API calls\n- Submit PR: 2-3 API calls\n- **Total: ~15+ API calls per workflow cycle**\n\n### With toolkit skills\n- Check reviews: 2-3 API calls\n- Address review: 2 API calls\n- Claim issue: 1 API call\n- Check workflow: 1 API call (GraphQL)\n- Submit PR: 1 API call\n- **Total: ~5-7 API calls per workflow cycle**\n\n**Result: 60-80% reduction in GitHub API usage**\n\n## Template Variables\n\nAll templates use `{{VARIABLE}}` syntax for placeholders:\n\n| Variable | Description | Example |\n|----------|-------------|---------|\n| `{{PROJECT_NAME}}` | Repository name | `my-project` |\n| `{{REPO_OWNER}}` | GitHub owner/org | `mycompany` |\n| `{{DEFAULT_BRANCH}}` | Main branch name | `main` |\n| `{{TEST_COMMAND}}` | How to run tests | `npm test` |\n| `{{TEST_UNIT_COMMAND}}` | Unit tests only | `npm run test:unit` |\n| `{{TEST_INTEGRATION_COMMAND}}` | Integration tests | `npm run test:integration` |\n| `{{LINT_COMMAND}}` | How to check style | `npm run lint` |\n| `{{PHASE_PREFIX}}` | Phase label prefix | `phase-` |\n| `{{SKILLS_DIR}}` | Skills directory | `.claude/skills` |\n| `{{CODEX_SKILLS_DIR}}` | Codex skills mirror | `.codex/skills` |\n| `{{COMMANDS_DIR}}` | Command docs directory | `.claude/commands` |\n| `{{DOCS_DIR}}` | Docs directory | `docs` |\n| `{{CODEX_BOT_USER}}` | Codex review bot username | `chatgpt-codex-connector[bot]` |\n\nKeep `{{SKILLS_DIR}}` as the canonical skill location and point `{{CODEX_SKILLS_DIR}}` at the same files (symlink recommended) so Claude and Codex use identical commands.\n\n## Repository Structure\n\n```\nclaude-workflow-toolkit/\n├── SKILL.md                    # Instructions for Claude/Codex agents\n├── README.md                   # This file (human documentation)\n├── LICENSE                     # MIT License\n│\n├── templates/\n│   ├── skills/                 # Workflow skill templates\n│   │   ├── check-reviews.sh.template     # Review detection\n│   │   ├── address-review.sh.template    # Review addressing\n│   │   ├── claim-issue.sh.template       # Issue claiming\n│   │   ├── check-workflow.sh.template    # Workflow validation\n│   │   ├── submit-pr.sh.template         # PR submission\n│   │   └── README.md.template\n│   │\n│   ├── .claude/\n│   │   ├── commands/           # Skill documentation templates\n│   │   │   ├── check-reviews.md.template\n│   │   │   ├── address-review.md.template\n│   │   │   ├── claim-issue.md.template\n│   │   │   ├── check-workflow.md.template\n│   │   │   └── submit-pr.md.template\n│   │   ├── SECURITY-CHECKLIST.md.template       # Generic\n│   │   ├── SECURITY-CHECKLIST-php.md.template   # PHP/WordPress\n│   │   ├── SECURITY-CHECKLIST-node.md.template  # Node.js\n│   │   ├── SECURITY-CHECKLIST-python.md.template # Python\n│   │   ├── SECURITY-CHECKLIST-bash.md.template  # Bash\n│   │   └── settings.local.json.template\n│   │\n│   └── docs/                   # Documentation templates\n│       ├── QUICK-REFERENCE.md.template\n│       ├── FAQ-AGENTS.md.template\n│       └── CODEBASE-MAP.md.template\n│\n├── profiles/                   # Pre-configured project settings\n│   ├── default.yaml\n│   ├── php-composer.yaml\n│   ├── bash-cli.yaml\n│   ├── node-npm.yaml\n│   └── python-poetry.yaml\n│\n├── examples/\n│   └── applied/               # Example generated output\n│\n└── scripts/\n    ├── setup-labels.sh        # Create required GitHub labels\n    └── validate-templates.sh  # Template syntax validator\n```\n\n## Setup Scripts\n\n### `scripts/setup-labels.sh`\n\nCreates the required GitHub labels in your repository:\n- **Workflow labels**: `agent-ready`, `in-progress`, `needs-review`, `blocked`\n- **Phase labels**: `phase-0` through `phase-6`\n- **Type labels**: `bug`, `enhancement`, `documentation`\n\n```bash\n./scripts/setup-labels.sh\n```\n\n## Why Use This?\n\n### Before (Manual Workflow)\n```bash\n# Did anyone review my PRs? Manual checking...\ngh pr list --state open\ngh pr view 42 --comments\n# Easy to miss reviews and start new work prematurely\n\n# Claiming an issue - 4 separate commands\ngh issue view 35\ngh issue edit 35 --remove-label \"agent-ready\"\ngh issue edit 35 --add-label \"in-progress\"\ngit checkout -b 35-feature-name\n\n# Each command = separate API call\n# More commands to remember\n# Easy to forget label updates\n```\n\n### After (Skill-Based Workflow)\n```bash\n# First: Check if any reviews need attention\n/check-reviews\n# Shows all PRs with unaddressed feedback, with priority levels\n\n# Address any reviews found\n/address-review 42\n# Checks out PR branch, shows all feedback organized by file\n\n# Then claim new work\n/claim-issue 35\n# Single API call, atomic operation, can't forget labels\n```\n\n### Key Benefits\n- **Review-first workflow**: Never start new work with unaddressed feedback\n- **Atomic operations**: Labels and branches updated together\n- **60-80% fewer API calls**: GraphQL and batching optimizations\n- **Security checklists**: Tech-specific guides for secure coding\n\n## Contributing\n\nContributions welcome! Please:\n1. Test templates against real projects\n2. Keep templates generic (no project-specific references)\n3. Document new variables in SKILL.md\n4. Add examples for new profiles\n\n### Adding a New Profile\n\n1. Create `profiles/your-stack.yaml`\n2. Define appropriate `test_command`, `lint_command`, etc.\n3. Add any stack-specific conventions\n4. Update README.md profiles table\n\n### Improving Templates\n\n1. Templates should work after variable substitution\n2. Use `{{VARIABLE}}` syntax consistently\n3. Test by mentally substituting real values\n4. Include comments explaining key logic\n\n## License\n\nMIT License - See [LICENSE](LICENSE)\n\n## Related Projects\n\n- [Claude Code](https://claude.ai/claude-code) - Anthropic's official CLI for Claude\n- Codex agents - Mirror `.claude/skills` to `.codex/skills` so Codex runs the same skills\n- [GitHub CLI](https://cli.github.com/) - GitHub's official command line tool\n\n---\n\n**Created to reduce token usage and improve workflow consistency for Claude Code and Codex agents.**\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fait88%2Fclaude-workflow-toolkit","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fait88%2Fclaude-workflow-toolkit","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fait88%2Fclaude-workflow-toolkit/lists"}