Skip to content
Back to search

openagentskill

The skill layer for AI agents: npm for AI Agent Skills.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/Leon-Drq/openagentskill with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/Leon-Drq/openagentskill'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/leon-drq--openagentskill/badge.svg)](/listings/leon-drq--openagentskill)

HTML

<a href="/listings/leon-drq--openagentskill"><img src="/api/listings/leon-drq--openagentskill/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Run the pinned, dependency-free GitHub Release:

npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz resolve "extract tables from PDF reports" --agent codex

Useful commands:

# Search without installing
npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz find "browser automation"

# Inspect a reviewed install plan
npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz add <skill-slug> --agent codex --dry-run

# Read the outcome reporting contract
npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz outcome-contract

The CLI refuses blocked Skills, requires explicit confirmation before execution, and supports `--no-telemetry`. Source and usage details live in [`packages/cli`](./packages/cli/README.md).

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 248 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <div align="center">
  2. <img src="./public/openagentskill-logo.svg" alt="OpenAgentSkill" width="300" />
  3. # OpenAgentSkill
  4. **The skill layer for AI agents.**
  5. Find, compare, audit, and install the right reusable Agent Skill before an agent acts.
  6. [![CI](https://img.shields.io/github/actions/workflow/status/Leon-Drq/openagentskill/ci.yml?branch=main&style=flat-square&label=CI)](https://github.com/Leon-Drq/openagentskill/actions/workflows/ci.yml)
  7. [![GitHub Release](https://img.shields.io/github/v/release/Leon-Drq/openagentskill?display_name=tag&style=flat-square)](https://github.com/Leon-Drq/openagentskill/releases)
  8. [![GitHub Stars](https://img.shields.io/github/stars/Leon-Drq/openagentskill?style=flat-square)](https://github.com/Leon-Drq/openagentskill/stargazers)
  9. [![License](https://img.shields.io/github/license/Leon-Drq/openagentskill?style=flat-square)](./LICENSE)
  10. [**Try Resolve →**](https://www.openagentskill.com/resolve) · [**Install the CLI →**](#install-the-cli)
  11. <br />
  12. <img src="./public/github-homepage-preview.png" alt="OpenAgentSkill task-to-skill resolver and registry" width="920" />
  13. </div>
  14. ## Why OpenAgentSkill?
  15. Agent Skills are easy to publish and increasingly hard to evaluate. A repository can be popular, recently updated, or well documented without being the right choice for a specific agent task.
  16. OpenAgentSkill adds the decision layer between discovery and execution:
  17. | Capability | What an agent gets |
  18. | --- | --- |
  19. | Task-to-skill resolve | One recommended Skill plus relevant alternatives |
  20. | Trust and audit signals | License, maintenance, install safety, permission, and quality context |
  21. | Install receipt | A stable, target-specific handoff for Codex, Claude Code, Cursor, or CLI |
  22. | Outcome loop | Evidence from successful, failed, blocked, or setup-required agent runs |
  23. | Machine-readable surfaces | JSON, text, OpenAPI, manifests, rankings, and public research datasets |
  24. OpenAgentSkill is not a static list and does not claim third-party code is safe. It helps agents make a better, inspectable decision before installation.
  25. ## 30-second demo
  26. Resolve a real task:
  27. ```bash
  28. curl "https://www.openagentskill.com/api/agent/resolve?task=extract+tables+from+PDF+reports&agent=codex&max_risk=medium&format=text"
  29. ```
  30. Example response:
  31. ```text
  32. OpenAgentSkill Resolve
  33. Task: extract tables from PDF reports
  34. Best skill: <recommended skill>
  35. Trust Score v5: <score and decision>
  36. Install: <reviewed install command>
  37. Risk: <risk level>
  38. Alternatives: <ranked alternatives>
  39. Receipt: <stable install handoff>
  40. ```
  41. The same resolver is available as [JSON](https://www.openagentskill.com/api/agent/resolve?task=extract+tables+from+PDF+reports&agent=codex&max_risk=medium), [plain text](https://www.openagentskill.com/api/agent/resolve?task=extract+tables+from+PDF+reports&agent=codex&max_risk=medium&format=text), and a compact lockfile format.
  42. ## Install the CLI
  43. Run the pinned, dependency-free GitHub Release:
  44. ```bash
  45. npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz resolve "extract tables from PDF reports" --agent codex
  46. ```
  47. Useful commands:
  48. ```bash
  49. # Search without installing
  50. npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz find "browser automation"
  51. # Inspect a reviewed install plan
  52. npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz add <skill-slug> --agent codex --dry-run
  53. # Read the outcome reporting contract
  54. npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.3.0/openagentskill-0.3.0.tgz outcome-contract
  55. ```
  56. The CLI refuses blocked Skills, requires explicit confirmation before execution, and supports `--no-telemetry`. Source and usage details live in [`packages/cli`](./packages/cli/README.md).
  57. ## Give this to an agent
  58. ```text
  59. Before installing a third-party Skill, call:
  60. GET https://www.openagentskill.com/api/agent/resolve
  61. ?task={TASK}
  62. &agent={AGENT}
  63. &max_risk=medium
  64. &format=json
  65. Inspect the recommended skill, alternatives, Trust Score v5, audit URL,
  66. risk level, install command, and install receipt. Use a sandbox first.
  67. After one narrow run, report the outcome to:
  68. POST https://www.openagentskill.com/api/agent/outcome
  69. ```
  70. Copy-ready templates for Codex, Claude Code, and Cursor are available in the [Agent Integration Kit](https://www.openagentskill.com/agent/integration-kit).
  71. ## How it works
  72. ```mermaid
  73. flowchart LR
  74. A["Agent task"] --> B["Resolve API"]
  75. B --> C["Candidate retrieval"]
  76. C --> D["Task fit + Trust Score v5"]
  77. D --> E["Best Skill + alternatives"]
  78. E --> F["Audit + install receipt"]
  79. F --> G["Sandboxed agent run"]
  80. G --> H["Outcome feedback"]
  81. H --> D
  82. ```
  83. Ranking combines task relevance with repository evidence, install readiness, maintenance, license clarity, risk signals, and real agent outcomes. See [Resolve Evals](https://www.openagentskill.com/evals/resolve) for the public evaluation surface.
  84. ## Core interfaces
  85. | Interface | Purpose |
  86. | --- | --- |
  87. | [Resolve Workbench](https://www.openagentskill.com/resolve) | Turn a task into a recommended Skill and install plan |
  88. | [Skill Registry](https://www.openagentskill.com/skills) | Search and filter indexed Skills |
  89. | [Rankings](https://www.openagentskill.com/rankings) | Compare trending, trusted, and Agent-Proven Skills |
  90. | [Audit Index](https://www.openagentskill.com/audits) | Inspect install, maintenance, license, and risk evidence |
  91. | [Skill Packs](https://www.openagentskill.com/skill-packs) | Compose reviewed Skills into reusable workflows |
  92. | [Creator Kit](https://www.openagentskill.com/creator-kit) | Submit, claim, badge, and share a Skill |
  93. | [API Docs](https://www.openagentskill.com/api-docs) | Integrate registry and agent endpoints |
  94. Machine-readable entry points:
  95. - [`llms.txt`](https://www.openagentskill.com/llms.txt)
  96. - [Agent manifest](https://www.openagentskill.com/.well-known/agent-manifest.json)
  97. - [OpenAgentSkill identity](https://www.openagentskill.com/.well-known/openagentskill.json)
  98. - [OpenAPI document](https://www.openagentskill.com/openapi.json)
  99. - [State of Agent Skills dataset](https://www.openagentskill.com/reports/state-of-agent-skills-2026.json)
  100. ## TypeScript SDK
  101. The dependency-free SDK source lives in [`packages/sdk`](./packages/sdk/README.md):
  102. ```ts
  103. import { OpenAgentSkill } from './packages/sdk/openagentskill.mjs'
  104. const client = new OpenAgentSkill({
  105. baseUrl: 'https://www.openagentskill.com',
  106. })
  107. const plan = await client.resolve('audit a repository', {
  108. agent: 'codex',
  109. maxRisk: 'medium',
  110. })
  111. await client.reportOutcome({
  112. event_id: plan.feedback.event_id,
  113. skill_slug: plan.recommendation.best_skill.slug,
  114. task: plan.task,
  115. outcome: 'success',
  116. dry_run: true,
  117. })
  118. ```
  119. The SDK package is prepared for npm publication but is not presented here as published until the public registry release is verifiable.
  120. ## For Skill authors
  121. Paste a repository, subdirectory, or `SKILL.md` URL into [Submit Skill](https://www.openagentskill.com/submit). Submissions are saved first and reviewed asynchronously; GitHub stars, a predefined category, and perfect metadata are not required.
  122. After indexing, authors receive:
  123. - A canonical Skill page and machine-readable metadata.
  124. - Trust, audit, quality, and Agent-Proven badge endpoints.
  125. - Claim and verified-maintainer paths.
  126. - Alternatives and use-case pages that can send qualified traffic back to the source.
  127. Browse the curated [GitHub Skill Index](./skills/README.md) for domain and scenario maps.
  128. ## Trust and security
  129. OpenAgentSkill never treats popularity as proof of safety. Trust Score and audits are decision-support signals, not certifications or guarantees.
  130. Before executing third-party code:
  131. 1. Read the source and install scripts.
  132. 2. Review permissions, network calls, dependencies, and required secrets.
  133. 3. Start in a sandbox or low-risk workspace.
  134. 4. Keep production credentials and customer data out of unreviewed Skills.
  135. Report vulnerabilities privately through the repository's **Security** tab. Report risky or misleading third-party listings through the [Skill data issue form](https://github.com/Leon-Drq/openagentskill/issues/new/choose). See [SECURITY.md](./SECURITY.md) for scope and disclosure rules.
  136. ## Local development
  137. Requirements: Node.js 22 and pnpm 10.28.2.
  138. ```bash
  139. git clone https://github.com/Leon-Drq/openagentskill.git
  140. cd openagentskill
  141. corepack enable
  142. pnpm install
  143. cp .env.example .env.local
  144. pnpm dev
  145. ```
  146. Quality checks:
  147. ```bash
  148. pnpm run lint
  149. pnpm run typecheck
  150. pnpm test
  151. pnpm run build
  152. ```
  153. The public website requires Supabase configuration. Pure parsing, resolver-contract, SDK, CLI, and circuit-breaker regression tests run without production credentials.
  154. GitHub sign-in appears on both login and registration when `NEXT_PUBLIC_GITHUB_OAUTH_ENABLED=true` is set in the local environment (as in `.env.example`) and the GitHub provider is enabled in Supabase. Public environment variables are included at build time; restart development after changing the flag and rebuild before deploying a changed value.
  155. ## Repository map
  156. ```text
  157. app/ Next.js pages, APIs, manifests, reports, and agent surfaces
  158. components/ Product UI and client interactions
  159. lib/ Resolve, ranking, audit, indexer, database, and growth logic
  160. packages/ Publishable CLI and SDK packages
  161. scripts/ Regression tests, migrations, importers, and maintenance jobs
  162. skills/ Curated GitHub Skill index by domain and scenario
  163. supabase/ Database functions and migrations
  164. ```
  165. ## Contributing
  166. Contributions are welcome across product, APIs, indexing, trust, documentation, and Skill metadata.
  167. - Read [CONTRIBUTING.md](./CONTRIBUTING.md).
  168. - Look for [`good first issue`](https://github.com/Leon-Drq/openagentskill/labels/good%20first%20issue) or [`help wanted`](https://github.com/Leon-Drq/openagentskill/labels/help%20wanted).
  169. - Ask usage questions in [GitHub Discussions](https://github.com/Leon-Drq/openagentskill/discussions).
  170. - Use [SUPPORT.md](./SUPPORT.md) to choose the right support channel.
  171. Project decisions and maintainer responsibilities are documented in [GOVERNANCE.md](./GOVERNANCE.md). Releases are tracked in [CHANGELOG.md](./CHANGELOG.md), and planned work lives in [ROADMAP.md](./ROADMAP.md).
  172. The two-layer discovery, validation, deduplication, fast-track, and GitHub rate-safety design is documented in [docs/candidate-intake-pipeline.md](./docs/candidate-intake-pipeline.md).
  173. ## License
  174. [MIT](./LICENSE) © OpenAgentSkill contributors.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.