Grades are a hygiene check, not a security guarantee.

Claude Code, Skill

Agent Skill Kit

Write an agent skill once.

Compatibility: Not stated.

Grade B

Minor issues only, no Medium or High. Worth a quick look at the findings below before installing.

Graded by PlugVetRubric v2

Not a guarantee. A hygiene check of the source at the commit shown, not a security audit.

PERMISSION_DELETE_FILES INSTALL_UNPINNED NETWORK_NO_FIXED_HOST SECRETS_PATH_DOTENV

Why this grade

  • PERMISSION_DELETE_FILES Low

    Contains a call that can delete files.

    agents/base.py, line 188 shutil.rmtree(target)

  • INSTALL_UNPINNED Low

    Dependencies use a range or git URL with no lockfile.

    requirements.txt, line 1 click>=8.0.0

  • NETWORK_NO_FIXED_HOST Low

    Makes a network call whose host is not written as a fixed address.

    scripts/update_homebrew.py, line 54 with urllib.request.urlopen(url, timeout=30) as response:

  • SECRETS_PATH_DOTENV Low

    Mentions a .env file.

    skills/coding/ask-docker-expert/README.md, line 40 * Use `.env` files for secrets (never hardcode passwords).

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/base.py, line 188

    shutil.rmtree(target)

    agents/antigravity/adapter.py, line 99

    shutil.rmtree(dst)

    agents/claude/adapter.py, line 126

    shutil.rmtree(target_item)

    Show more

    agents/codex/adapter.py, line 101

    shutil.rmtree(dst)

    agents/cursor/adapter.py, line 106

    shutil.rmtree(dst)

    agents/gemini/adapter.py, line 89

    shutil.rmtree(dst)

    agents/universal/adapter.py, line 83

    shutil.rmtree(dst)

    agents/universal/adapter.py, line 88

    shutil.rmtree(dst)

    ask/utils/git.py, line 42

    subprocess.run(["git", "pull"], cwd=str(target_dir), check=True, capture_output=True)

    ask/utils/git.py, line 48

    subprocess.run(["git", "clone", url, str(target_dir)], check=True, capture_output=True)

    scripts/update_homebrew.py, line 118

    subprocess.run(["git", "-C", str(tap_path), "add", "."], check=True)

    scripts/update_homebrew.py, line 123

    staged = subprocess.run(

    scripts/update_homebrew.py, line 130

    subprocess.run(["git", "-C", str(tap_path), "commit", "-m", f"Update agent-skill-kit to v{version}"], check=True)

    scripts/update_homebrew.py, line 131

    subprocess.run(["git", "-C", str(tap_path), "push"], check=True)

  • Install and update

    Concern

    dependencies are not pinned.

    INSTALL_UNPINNED

    requirements.txt, line 1

    click>=8.0.0

  • Network

    Concern

    can reach a host that is not written as a fixed address.

    NETWORK_NO_FIXED_HOST

    scripts/update_homebrew.py, line 54

    with urllib.request.urlopen(url, timeout=30) as response:

    skills/coding/ask-explaining-code/README.md, line 244

    > return requests.get(url)

    skills/coding/ask-vue-architect/README.md, line 43

    * ❌ **Avoid:** Manually creating `ref` for each field and calling `axios.post`.

    Show more

    skills/coding/ask-vue-architect/README.md, line 84

    * **Rule:** If the `ziggy-js` library is detected (standard in Laravel), NEVER hardcode URLs like `axios.post('/api/user')`.

    skills/coding/ask-vue-architect/SKILL.md, line 9

    ❌ NO axios.post for forms → use `useForm` from Inertia

  • Secrets

    Concern

    reads a credential or environment file.

    SECRETS_PATH_DOTENV

    .gitignore, line 53

    # Environments

    skills/coding/ask-docker-expert/README.md, line 40

    * Use `.env` files for secrets (never hardcode passwords).

    skills/coding/ask-docker-expert/SKILL.md, line 10

    ❌ NO hardcoded secrets → use .env files

    Show more

    skills/coding/ask-docker-expert/SKILL.md, line 44

    - Use .env for secrets

    skills/coding/ask-security-sentinel/README.md, line 13

    * **Action:** If found, HALT immediately and warn the user to move it to `.env`.

    skills/coding/ask-security-sentinel/SKILL.md, line 17

    → If found: HALT, warn user, move to .env

    skills/tooling/ask-system-architect-prime/README.md, line 46

    3. Locate configuration sources (package.json, pyproject.toml, .env.example)

    skills/tooling/ask-wiki-init/SKILL.md, line 43

    1. **Analyze Codebase**: Read `README.md` and *any* other `.md` files in the root. You MUST also actively read `composer.json`, `package.json`, `docker-compose.yml`, `.env.example`

  • 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.

Get the full report on this tool, £49

Note

Hygiene check of commit 760c679d6f84. can delete files. dependencies are not pinned. can reach a host that is not written as a fixed address. reads a credential or environment file.

Date and source

Date checked:

View source

github.com

Commit graded: 760c679d6f84cf7fe54471a9acb400b9d1ee936f

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.

Read how this letter is chosen