A collection of Agent Skills — portable, version-controlled procedural knowledge for AI coding agents, following the open Agent Skills spec.
Current version: 0.18.0 (see CHANGELOG.md).
Documentation: devopam.github.io/agent-skills · Agent index: llms.txt
| Skill | What it does | Use when |
|---|---|---|
project-incubation |
Repo structure, architecture, tech-stack baseline at inception and re-audit. | New or existing repos vs incubation baseline. |
python-code-review |
Python review across 11 domains with a scored report. | Quality/security/production readiness. |
typescript-code-review |
TypeScript review across 11 domains with a scored report. | TS libraries, apps, shared packages. |
nodejs-code-review |
Node.js service/API review across 11 domains. | Express/Fastify/Hono/Nest, workers, CLIs. |
react-code-review |
React/Next UI review across 11 domains. | Components, hooks, client security/perf. |
ci-cd-plumber |
CI/CD scaffold and audit with scored domains. | Pipelines, release automation. |
pr-review |
Pre-submit / PR readiness. | Before or during PR review. |
postgresql-review |
Live PostgreSQL review after MCPg readiness. | Postgres health, schema, security. |
ui-system-review |
UI system consistency (Web / Apple / Android). | Design-system drift, form factors. |
regulatory-compliance-applicability-scan |
Privacy/domain applicability scan (global packs). Suggests, does not certify. Not legal advice. | Multi-market privacy gap orientation. |
Point any agentskills.io-compliant agent at this repo (or copy skills/<name>/).
.claude-plugin/plugin.json — validate with
claude plugin validate .. Plugin version tracks 0.18.0.
See evals/README.md.
Contributor Covenant v3.0 — devopam@gmail.com or GitHub Issue.
MIT.