Skip to content
Back to search

skillproof

Test, score and showcase AI agent skills and prompts. Free, open source, browser-based.

Safety scan

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

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/piyush2707--skillproof/badge.svg)](/listings/piyush2707--skillproof)

HTML

<a href="/listings/piyush2707--skillproof"><img src="/api/listings/piyush2707--skillproof/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

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 · 256 lines
  1. <div align="center">
  2. <img src="assets/skill-logo.jpg" alt="SkillProof: agent skills are easy to write, hard to trust" width="100%">
  3. <br>
  4. **Free, open-source tools to check, measure and showcase AI agent skills.**<br>
  5. Everything runs in your browser. Nothing is uploaded.
  6. <br>
  7. [![Validate](https://github.com/piyush2707/skillproof/actions/workflows/validate.yml/badge.svg)](https://github.com/piyush2707/skillproof/actions/workflows/validate.yml)
  8. ![License: MIT](https://img.shields.io/badge/license-MIT-d99a55?style=flat-square)
  9. ![Status](https://img.shields.io/badge/status-early%20development-5a3a24?style=flat-square)
  10. ![No install](https://img.shields.io/badge/install-none-14100d?style=flat-square)
  11. [![Stars](https://img.shields.io/github/stars/piyush2707/skillproof?style=flat-square&color=d99a55)](https://github.com/piyush2707/skillproof/stargazers)
  12. <br>
  13. ### [Open the live site](https://piyush2707.github.io/skillproof/)
  14. [Context Meter](https://piyush2707.github.io/skillproof/meter/) ·
  15. [Conflict Map](https://piyush2707.github.io/skillproof/conflicts/) ·
  16. [Eval Cards](https://piyush2707.github.io/skillproof/evals/) ·
  17. [Skill Pages](https://piyush2707.github.io/skillproof/skills/)
  18. </div>
  19. <br>
  20. ---
  21. ## The problem
  22. A skill is a folder with a `SKILL.md` file that teaches an agent one task. Skills are quick to write, but it is hard to know:
  23. - Is it **well-formed**?
  24. - How much **context** does it use before any task starts?
  25. - Does it **clash** with the other skills you installed?
  26. - Is there any **evidence** that it actually helps?
  27. Most skill repositories are collections of instructions with no proof. SkillProof adds measurable checks and visible evidence, using only GitHub and your browser.
  28. ---
  29. ## The toolkit
  30. <table>
  31. <tr>
  32. <td width="50%" valign="top">
  33. ### Context Meter
  34. Estimate how many tokens your skills use before any task starts.
  35. [**Open →**](https://piyush2707.github.io/skillproof/meter/)
  36. </td>
  37. <td width="50%" valign="top">
  38. ### Conflict Map
  39. Find skills whose descriptions overlap and may confuse an agent.
  40. [**Open →**](https://piyush2707.github.io/skillproof/conflicts/)
  41. </td>
  42. </tr>
  43. <tr>
  44. <td width="50%" valign="top">
  45. ### Eval Cards
  46. A JSON format and viewer for with/without-skill test results.
  47. [**Open →**](https://piyush2707.github.io/skillproof/evals/)
  48. </td>
  49. <td width="50%" valign="top">
  50. ### Skill Pages
  51. A page per example skill: cost, install steps and evidence.
  52. [**Open →**](https://piyush2707.github.io/skillproof/skills/)
  53. </td>
  54. </tr>
  55. <tr>
  56. <td colspan="2" valign="top">
  57. ### Validator Action
  58. A GitHub Action that checks `SKILL.md` files and eval cards on every push and pull request. It also checks that deliberately broken fixtures are rejected, so the validator itself is tested. [**View the code →**](validators/)
  59. </td>
  60. </tr>
  61. </table>
  62. ---
  63. ## Try it in 30 seconds
  64. No install. No account.
  65. 1. Open the [Context Meter](https://piyush2707.github.io/skillproof/meter/) and tap **Load sample**.
  66. 2. Paste your own `SKILL.md` and tap **Measure**.
  67. 3. Paste several skills, separated by a line of `=====`, into the [Conflict Map](https://piyush2707.github.io/skillproof/conflicts/).
  68. 4. Record test results in an [Eval Card](schema/eval-card.md) and view them on the [Eval Cards page](https://piyush2707.github.io/skillproof/evals/).
  69. ---
  70. ## How it fits together
  71. ```mermaid
  72. flowchart LR
  73. A[SKILL.md] --> B[Context Meter]
  74. A --> C[Conflict Map]
  75. A --> D[Skill Pages]
  76. E[Eval Card JSON] --> F[Eval Cards viewer]
  77. E --> D
  78. G[Pull request] --> H[Validator Action]
  79. A --> H
  80. E --> H
  81. ```
  82. Everything on the left is plain text in the repository. Everything on the right is either a web page in your browser or a GitHub Action.
  83. ---
  84. ## What the validator checks
  85. The workflow in [`.github/workflows/validate.yml`](.github/workflows/validate.yml) runs on every push and pull request.
  86. | Check | Looks at |
  87. |---|---|
  88. | Frontmatter | Exists, with a `name` and `description` |
  89. | Name | Lowercase letters, numbers and hyphens only |
  90. | Description | Long enough to be useful and says when to use the skill |
  91. | File links | Links inside the skill point to files that exist |
  92. | Length | Warns on very long skill bodies |
  93. | Eval cards | Required fields, valid results and unique case ids |
  94. Skill rules live in [`validators/rules.json`](validators/rules.json).
  95. ---
  96. ## Eval Cards
  97. An Eval Card is a file named `<skill>.eval.json` in [`evals/`](evals/). Each case records a prompt, what a good answer needs, and the result **without** and **with** the skill.
  98. | Result | Meaning |
  99. |---|---|
  100. | `pass` | The answer met the expectation |
  101. | `partial` | Partly met |
  102. | `fail` | Did not meet it |
  103. | `untested` | Not run |
  104. The `evidence_type` field says whether a card is `example`, `manual` or `automated`, so sample data is never mistaken for a real test. See the [format reference](schema/eval-card.md).
  105. ---
  106. ## Repository layout
  107. ```
  108. skillproof/
  109. ├── docs/ Website (GitHub Pages)
  110. │ ├── index.html Home page
  111. │ ├── meter/ Context Meter
  112. │ ├── conflicts/ Conflict Map
  113. │ ├── evals/ Eval Cards viewer
  114. │ ├── skills/ Skill Pages
  115. │ └── assets/ Banner image
  116. ├── skills-examples/ Example skills
  117. ├── evals/ Eval cards
  118. ├── schema/ Eval Card format reference
  119. ├── validators/ Validator scripts and rules
  120. ├── tests/fixtures/ Broken files the validator must reject
  121. └── .github/ Workflows and issue templates
  122. ```
  123. ---
  124. ## Honest limits
  125. - **Token counts are estimates**, at about 4 characters per token. A real tokenizer will differ.
  126. - **Conflict scores are word overlap.** They are a warning to review, not proof that an agent will get confused.
  127. - **SkillProof does not run any agent.** Eval Cards record evidence that a person ran and logged.
  128. - **Compatibility:** SkillProof reads the `SKILL.md` format (name and description in frontmatter). Support for specific agents will be listed here only after it has been tested.
  129. - The example eval card for `hello-skill` is sample data and is labelled that way.
  130. ---
  131. ## Roadmap
  132. - [x] Validator Action
  133. - [x] Context Meter
  134. - [x] Conflict Map
  135. - [x] Eval Cards
  136. - [x] Skill Pages
  137. - [ ] RAG Failure Attribution skill
  138. - [ ] RepoReel: turn a GitHub repo into a short explainer video in the browser
  139. - [ ] More validator checks and community-submitted skills
  140. Full details in [`ROADMAP.md`](ROADMAP.md).
  141. ---
  142. ## FAQ
  143. <details>
  144. <summary><b>Does SkillProof upload my skills anywhere?</b></summary>
  145. <br>
  146. No. The web tools run entirely in your browser. What you paste stays on your device.
  147. </details>
  148. <details>
  149. <summary><b>Does it work with my agent?</b></summary>
  150. <br>
  151. SkillProof reads the <code>SKILL.md</code> format. It does not run any agent, so it checks your files rather than an agent's behavior. Agent-specific compatibility will be listed only after testing.
  152. </details>
  153. <details>
  154. <summary><b>Why are token counts approximate?</b></summary>
  155. <br>
  156. They use about 4 characters per token, which is a rough rule. Different models count tokens differently, so treat the numbers as a guide for comparing skills.
  157. </details>
  158. <details>
  159. <summary><b>Can I add my own skill?</b></summary>
  160. <br>
  161. Yes. Add a folder under <code>skills-examples/</code> with a <code>SKILL.md</code> and open a pull request. See <a href="CONTRIBUTING.md">CONTRIBUTING.md</a>.
  162. </details>
  163. <details>
  164. <summary><b>Can I trust the example eval card?</b></summary>
  165. <br>
  166. It is labelled "example" because it is sample data showing the format, not a real test run. Real results use <code>manual</code> or <code>automated</code>.
  167. </details>
  168. ---
  169. ## Contributing
  170. Skills, eval cards and fixes are welcome. Read [`CONTRIBUTING.md`](CONTRIBUTING.md) first and follow the [Code of Conduct](CODE_OF_CONDUCT.md). Report vulnerabilities privately as described in [`SECURITY.md`](SECURITY.md).
  171. If SkillProof is useful to you, a star on the repository helps others find it.
  172. ---
  173. ## License
  174. MIT. See [`LICENSE`](LICENSE).
  175. <br>
  176. <div align="center">
  177. Made by [Piyush](https://github.com/piyush2707) at **Welvox**
  178. </div>

Ratings and comments

Sign in to rate this skill.

Sign in

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