Claude Code, Skill
Agent Skills
Reusable agent skills for AI coding assistants.
Compatibility: Not stated.
Grade C
One Medium finding. Read the report before installing.
Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.
SECRETS_PATH_CREDENTIAL
Why this grade
- SECRETS_PATH_CREDENTIAL Medium
Mentions a credential store path (~/.ssh, id_rsa, .npmrc, .aws/credentials, credentials.json, .netrc).
benchmarks/app-ai-guardrails/fixtures/make-golden.sh, line 252
printf 'min-release-age=7\n' > .npmrc
Any High, or two or more Medium findings, give D. One Medium gives C. Low findings or a security Note give B. Each code names a finding behind the letter.
Grades are a hygiene check of the source at the commit shown, on the date checked, not a security audit or a guarantee. The free directory grades are automated. Before release, a PlugVet engineer reviews every paid report and any grade we hold back, including all F grades, against the source code and the evidence we quote.
What was checked
Plain notes from the hygiene checks.
Permissions
Concern
can delete files.
PERMISSION_DELETE_FILES
.agents/skills/skill-creator/eval-viewer/generate_review.py, line 291
result = subprocess.run(
.agents/skills/skill-creator/scripts/improve_description.py, line 35
result = subprocess.run(
.agents/skills/skill-creator/scripts/run_eval.py, line 85
process = subprocess.Popen(
Show more
benchmarks/app-ai-guardrails/check-guardrails.mjs, line 208
fs.rmSync(tmpRoot, { recursive: true, force: true });
benchmarks/app-ai-guardrails/check-guardrails.mjs, line 232
fs.rmSync(path.join(tmpRoot, rel), { force: true });
benchmarks/app-ai-guardrails/check-guardrails.mjs, line 239
fs.rmSync(tmpRoot, { recursive: true, force: true });
benchmarks/database-documentation/evaluation/selftest.mjs, line 25
process.on('exit', () => fs.rmSync(work, { recursive: true, force: true }));
benchmarks/production-grade/score.js, line 90
const f = tmp('.py', harness); const r = run(`"${py()}" "${f}"`); try { fs.unlinkSync(f); } catch (e) {}
benchmarks/production-grade/score.js, line 108
const f = tmp('.py', harness); const r = run(`"${py()}" "${f}"`); try { fs.unlinkSync(f); } catch (e) {}
benchmarks/production-grade/score.js, line 126
const f = tmp('.ts', harness); const r = run(`npx --no-install tsx "${f}"`); try { fs.unlinkSync(f); } catch (e) {}
Install and update
Pass
No install-time script found, and dependencies are pinned or absent.
Network
Concern
can reach a host that is not written as a fixed address. can reach start.spring.io. The description does not name that host. can reach api.anthropic.com. The description does not name that host.
NETWORK_NO_FIXED_HOST NETWORK_UNNAMED_HOST
benchmarks/app-ai-guardrails/fixtures/make-bare.sh, line 9
# `next` and `springboot` are implemented. `springboot` is trivial (one curl + tar + git init, no
benchmarks/app-ai-guardrails/fixtures/make-bare.sh, line 41
curl -sS https://start.spring.io/starter.tgz \
benchmarks/create-skill-autoresearch/scoring/score.mjs, line 178
const res = await fetch("https://api.anthropic.com/v1/messages", {
Show more
skills/app-ai-guardrails/references/adapters/springboot.md, line 54
curl -sS https://start.spring.io/starter.tgz \
skills/create-skill-autoresearch/references/pipeline-phases.md, line 195
# Step 3-4: Invoke skill via LLM API (curl to OpenAI-compatible endpoint).
skills/create-skill-autoresearch/references/pipeline-phases.md, line 207
SKILL_OUTPUT=$(curl -s "${JUDGE_API_BASE}/chat/completions" \
.agents/skills/skill-creator/eval-viewer/viewer.html, line 666
const resp = await fetch("/api/feedback");
.agents/skills/skill-creator/eval-viewer/viewer.html, line 1012
fetch("/api/feedback", {
Secrets
Concern
reads a credential or environment file.
SECRETS_ENV_READ SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV
benchmarks/app-ai-guardrails/fixtures/make-golden.sh, line 252
printf 'min-release-age=7\n' > .npmrc
benchmarks/create-skill-autoresearch/scoring/score.mjs, line 234
let oauthToken = process.env.CLAUDE_CODE_OAUTH_TOKEN;
benchmarks/create-skill-autoresearch/scoring/score.mjs, line 237
const report = await scoreSkill(skillDir, { apiKey: process.env.ANTHROPIC_API_KEY, oauthToken, executorOut });
Show more
benchmarks/database-documentation/evaluation/evaluate.sh, line 10
# target-id -> targets/<target-id>.env (DSN, dialect, schemas)
skills/production-grade/references/09-before-after.md, line 230
// 3. Secrets hygiene - .env.example (or config.example.yaml, etc.) checked in with
.gitignore, line 3
*.log
AGENTS.md, line 77
implementation. Never commit secrets; `.env` is git-ignored.
benchmarks/production-grade/.gitignore, line 4
node_modules/
skills/database-documentation/SKILL.md, line 55
and connection URLs in `.env*`.
skills/database-documentation/SKILL.md, line 59
`.env`. **Prove it** with one trivial query (`SELECT 1`, list tables) before continuing. If the DB is
Agent instructions
Pass
No instruction to ignore earlier instructions or to hide a step.
Obfuscation
Pass
No encoded or minified payload found.
Licence
Pass
A licence file is present (LICENSE).
Maintainer activity
Pass
The latest commit is within the last 365 days.
Note
Hygiene check of commit 99652ee6a787. can delete files. can reach a host that is not written as a fixed address. can reach start.spring.io. The description does not name that host. can reach api.anthropic.com. The description does not name that host. reads a credential or environment file.
Date and source
Date checked:
github.com
Commit graded: 99652ee6a78759a5ef655388497a2b889ad6c602
This link goes to the original host. PlugVet does not host or check what you download from there, and installing it is your own decision and risk.