{"id":50804441,"url":"https://github.com/jimtin/production-ai","last_synced_at":"2026-06-12T23:35:14.644Z","repository":{"id":363935998,"uuid":"1265625453","full_name":"jimtin/production-ai","owner":"jimtin","description":"Skills and guardrails that make AI coding agents prove their work: planning gates, test preflights, fail-closed deployment, and a self-improving skill library.","archived":false,"fork":false,"pushed_at":"2026-06-11T06:07:12.000Z","size":321,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-06-12T23:35:01.970Z","etag":null,"topics":["agent-skills","ai-agents","ci-cd","claude-code","codex-cli","guardrails"],"latest_commit_sha":null,"homepage":null,"language":"JavaScript","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/jimtin.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":"2026-06-11T00:14:21.000Z","updated_at":"2026-06-11T06:07:17.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/jimtin/production-ai","commit_stats":null,"previous_names":["jimtin/production-ai"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/jimtin/production-ai","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jimtin%2Fproduction-ai","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jimtin%2Fproduction-ai/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jimtin%2Fproduction-ai/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jimtin%2Fproduction-ai/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/jimtin","download_url":"https://codeload.github.com/jimtin/production-ai/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jimtin%2Fproduction-ai/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34266915,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-12T02:00:06.859Z","response_time":109,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"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":["agent-skills","ai-agents","ci-cd","claude-code","codex-cli","guardrails"],"created_at":"2026-06-12T23:35:13.965Z","updated_at":"2026-06-12T23:35:14.636Z","avatar_url":"https://github.com/jimtin.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# production-ai\n\n**Skills and guardrails that make an AI coding agent prove its work.**\n\nThis is a working library, not a prompt collection. Every artifact here was extracted from a real daily-driver setup where AI agents plan features, write code, run containerized test gates, review pull requests, and promote releases to production — with humans setting policy and agents carrying the proof burden.\n\nThe core idea: coding agents optimize for *looks done*. Left alone, they will write plausible code, skip the boring verification, and report success. Everything in this repo exists to redefine \"done\" as **proven** — with tests, evidence, fail-closed gates, and explicit completion blockers the agent cannot talk its way past.\n\n## The library\n\nEach skill has two layers:\n\n- `skills/\u003cname\u003e/` — the clean installable payload loaded by the agent when the skill triggers.\n- `docs/skills/\u003cname\u003e.md` — the public human guide: why the skill exists, the failure it prevents, and how to adapt it.\n\n| Skill | Kind | Guide | Payload |\n|---|---|---|---|\n| clarify-before-build | Planning gate | [guide](docs/skills/clarify-before-build.md) | [payload](skills/clarify-before-build/) |\n| feature-design-preflight | Design gate | [guide](docs/skills/feature-design-preflight.md) | [payload](skills/feature-design-preflight/) |\n| user-action-coverage-review | Coverage gate | [guide](docs/skills/user-action-coverage-review.md) | [payload](skills/user-action-coverage-review/) |\n| frontend-design-quality | Quality gate | [guide](docs/skills/frontend-design-quality.md) | [payload](skills/frontend-design-quality/) |\n| test-readiness-preflight | Validation preflight | [guide](docs/skills/test-readiness-preflight.md) | [payload](skills/test-readiness-preflight/) |\n| full-app-review | Review orchestrator | [guide](docs/skills/full-app-review.md) | [payload](skills/full-app-review/) |\n| codebase-prune-review | Removal gate | [guide](docs/skills/codebase-prune-review.md) | [payload](skills/codebase-prune-review/) |\n| error-logging-instrumentation | Observability review | [guide](docs/skills/error-logging-instrumentation.md) | [payload](skills/error-logging-instrumentation/) |\n| repo-technical-documentation | Documentation engine | [guide](docs/skills/repo-technical-documentation.md) | [payload](skills/repo-technical-documentation/) |\n| nextjs-vercel-analytics | Implementation guide | [guide](docs/skills/nextjs-vercel-analytics.md) | [payload](skills/nextjs-vercel-analytics/) |\n| pr-production-gate | Deployment gate | [guide](docs/skills/pr-production-gate.md) | [payload](skills/pr-production-gate/) |\n| security-threat-model | Security gate | [guide](docs/skills/security-threat-model.md) | [payload](skills/security-threat-model/) |\n| repo-testing-setup | Foundation gate | [guide](docs/skills/repo-testing-setup.md) | [payload](skills/repo-testing-setup/) |\n| laptop-currency-maintenance | Ops automation | [guide](docs/skills/laptop-currency-maintenance.md) | [payload](skills/laptop-currency-maintenance/) |\n\n## The patterns\n\nThe skills are instances of a small set of system-level patterns, documented in [docs/](docs/):\n\n- [Skill anatomy](docs/skill-anatomy.md) — the structure every skill follows, and why each part earns its place.\n- [The skill graph](docs/skill-graph.md) — skills that invoke skills: orchestrators, specialist gates, and composition rules.\n- [The workspace constitution](docs/workspace-constitution.md) — one `AGENTS.md` that governs every repo in a workspace ([template](templates/AGENTS-workspace-template.md)).\n- [PR production gate](docs/patterns/pr-production-gate.md) — a fail-closed, container-only review-and-deploy train owned by your own machine.\n- [The learning loop](docs/patterns/learning-loop.md) — a nightly automation that mines the agent's own sessions and proposes skill improvements, safely.\n- [Sync and backup](docs/patterns/sync-and-backup.md) — version-controlling your agent configuration with redaction and fail-closed pushes.\n\n## Quickstart\n\nInstall a skill into Codex CLI:\n\n```bash\ngit clone https://github.com/jimtin/production-ai.git\ncd production-ai\nscripts/install-skill.sh clarify-before-build\n```\n\nThen ask your agent to plan something substantial — the skill triggers on planning requests, or invoke it directly with `$clarify-before-build`.\n\nAdopt the workspace constitution:\n\n```bash\ncp production-ai/templates/AGENTS-workspace-template.md ~/workspace/AGENTS.md\n```\n\nThen edit it down: delete every rule you cannot or will not enforce. A constitution the gates don't back up is worse than no constitution — the agent learns the rules are decorative.\n\nValidate the library (used by CI on every push):\n\n```bash\n./scripts/validate.sh        # structure: frontmatter, references, docs, clean payloads\n./scripts/privacy-scan.sh    # denylist sweep + gitleaks, fail closed\n```\n\n## Portability\n\nThese artifacts are written for [OpenAI Codex CLI](https://github.com/openai/codex) (skills live in `~/.codex/skills/`, workspace rules in `AGENTS.md`, cross-references use `$skill-name`). The patterns port directly to other agent harnesses:\n\n- **Claude Code**: skills go to `~/.claude/skills/`, workspace rules go in `CLAUDE.md`, and `$skill-name` references become skill-name mentions or slash commands. The anatomy (frontmatter description as trigger, progressive-disclosure references, scripts with tests) is identical.\n- **Anything else**: the gates are prose contracts. Any harness that can load instructions conditionally can run them.\n\n## What is deliberately not here\n\nThis is the public, sanitized derivative of a private setup. Excluded on purpose:\n\n- Client-specific CI integrations and business-domain skills (a market-research/ROI reporting skill, client API contracts).\n- Cached data of any kind — API responses, market data, databases, fixtures with real-world records.\n- Session logs, learning-loop summaries, and automation state — even sanitized ones.\n- Real channel IDs, hostnames, repo names, and absolute paths — replaced with placeholders.\n\nThe privacy guarantee is structural, not a one-time cleanup: [CI runs a fail-closed denylist sweep plus gitleaks](.github/workflows/ci.yml) on every push, using the same approach as the private setup it came from. Promotion from the private library into this repo is always a manual, reviewed step — never automated.\n\n## Creating content from this repo\n\nThe [content map](docs/content/content-map.md) tracks which artifacts are ready to become posts, videos, or talks, with the hook for each. The [glossary](docs/content/glossary.md) keeps terminology consistent across pieces. Each [skill guide](docs/skills/) is structured to be the first draft of its own article.\n\n## License\n\n[MIT](LICENSE). Build on it, ship it, write about it — attribution appreciated.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjimtin%2Fproduction-ai","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjimtin%2Fproduction-ai","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjimtin%2Fproduction-ai/lists"}