agentleFS
Sign inStart free
patronus-protectpatronus-protect/patronus-security-cli/.claude/skills/qa-dx/SKILL.md

qa-dx

Skill2 starsChanged 2 months agoApache-2.0View on GitHub

Developer-experience QA for Patronus — check that docs.patronus.studio is current and understandable, that SDK quick starts and examples work, and list what is missing. Use as part of the QA run or when asked to review docs or SDK usability.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run node qa/sdk.mjs. It installs the *published* SDKs into empty directories and runs each README quick start against the QA key.

    line 12
  2. Distinctive:

    Read the site the way the persona would, using WebFetch or the built-in browser.

    line 16
  3. Distinctive:

    Links: no 404s or redirects to other products.

    line 27
  4. Very specific:

    Write ~/.patronus-qa/runs/<date>/dx.json in the report shape (see the qa-run skill).

    line 51

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Developer experience QA

Persona: a developer who has never used Patronus, has 15 minutes, and wants to scan untrusted text from their app. Judge everything from that seat.

1. Deterministic baseline

Run node qa/sdk.mjs. It installs the published SDKs into empty directories and runs each README quick start against the QA key. Its findings already cover installability, version drift and broken quick starts.

2. Docs walk-through (docs.patronus.studio)

Read the site the way the persona would, using WebFetch or the built-in browser. Start at the landing page, then:

  • Quick start
  • Authentication / API keys
  • Each SDK page (TypeScript, Python, Rust)
  • API reference
  • Runtime hooks / plugins (Claude, Codex, DeepSeek)
  • Pricing / limits

For every page, record:

  • Correctness: compare parameters, endpoints and examples against https://docs.patronus.studio/openapi.json, the SDK sources (sdk/typescript/src, sdk/python/src, crates/patronus-api-client/src) and patronus-security-scanner --help (plus integration --help). Note outdated names, removed flags, wrong defaults and wrong versions.
  • Runnable examples: copy each code snippet into a scratch directory and run it with PATRONUS_API_KEY from PATRONUS_QA_API_KEY. A snippet that fails as written is a finding.
  • Links: no 404s or redirects to other products.
  • Clarity score 1–5. 5 means the persona succeeds without guessing. Explain anything below 4 in one sentence.
  • Missing: what the persona needed and could not find. Typical gaps are an error-handling example, rate limits, scan_url / scan_mcp_server examples, a minimal hooks walkthrough, and troubleshooting for degraded / inactive hook messages.

3. In-repo docs and examples

Review these for the same criteria:

  • README.md
  • docs/*.md
  • the SDK READMEs and INSTALL.md files
  • examples/file-checker/

Simple examples should exist for:

  • text,
  • URL,
  • file upload,
  • MCP server scan,
  • handling a pending result,
  • handling errors and limits.

Each missing simple example is a docs finding with a proposed snippet in the suggestion.

4. Output

Write ~/.patronus-qa/runs/<date>/dx.json in the report shape (see the qa-run skill). Use metrics.pages = [{url, score, issues: n}], and give findings area docs, sdk or ux.

Fixes:

  • In-repo docs, READMEs and examples are draft-PR candidates.
  • Fixes to docs.patronus.studio are proposed issues for the control-plane repository only. Quote the exact wrong text and the proposed replacement.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free