Grades are a hygiene check, not a security guarantee.

MCP, MCP server

asamm

An OWASP SAMM Extension for AI-Driven Development

Compatibility: Not stated.

Grade C

One Medium finding. Read the findings before installing.

Graded by PlugVetRubric v2

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

Rescanned at commit 53fa38e78cbf, unchanged since the check below.

SECRETS_PATH_CREDENTIAL

Why this grade

  • SECRETS_PATH_CREDENTIAL Medium

    Mentions a credential store path (~/.ssh, id_rsa, .npmrc, .aws/credentials, credentials.json, .netrc).

    audit/samples/claude-code-zhet/aSAMM-audit-claude-code-2026-04-13-basic.html, line 167 <code>~/.ssh/id_rsa</code> [empirical], open network egress [empirical], sandbox explicitly disabled by user opt-in

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

    Note

    requests shell access. The manifest names this permission.

    PERMISSION_SHELL_DECLARED

    tools/build_pdf.py, line 46

    result = subprocess.run(

    tools/build_reviewer_pack.py, line 72

    result = subprocess.run(command, cwd=ROOT, text=True, capture_output=True)

    tools/check_secrets.py, line 17

    result = subprocess.run(

    Show more

    tools/check_secrets.py, line 48

    result = subprocess.run(command, cwd=ROOT, text=True, capture_output=True)

    tools/verify_consistency.py, line 47

    result = subprocess.run(command, cwd=ROOT, text=True, capture_output=True)

    audit/analysis-principles.md, line 360

    result = subprocess.run(['curl', '-s', 'https://example.com', '-o', '/dev/null', '-w', '%{http_code}'],

    audit/environment-adapters.md, line 112

    result = subprocess.run(['cat','/proc/self/status'], capture_output=True, text=True)

  • 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.

    NETWORK_NO_FIXED_HOST

    audit/environment-adapters.md, line 119

    urllib.request.urlopen('[APPROVED-PUBLIC-CANARY]', timeout=3)

  • Secrets

    Concern

    reads a credential or environment file.

    SECRETS_PATH_CREDENTIAL SECRETS_PATH_DOTENV

    audit/agent-environment-profile.md, line 91

    - credential bleed from local config, `.env`, keychains, SSH, or shell history

    audit/samples/SecOps/SecOps-asamm-audit-2026-04-15-public.html, line 292

    <td>Rotate the Engine key; move to <code>.env</code>; remove <code>ONOTOLLE_SEARCH_API_KEY</code> default from <code>app/config/settings.py:176</code>. The git-history leak is perm

    audit/samples/claude-code-zhet/aSAMM-audit-claude-code-2026-04-13-basic.html, line 167

    <code>~/.ssh/id_rsa</code> [empirical], open network egress [empirical], sandbox explicitly disabled by user opt-in

    Show more

    audit/samples/SecOps/assam_retro_en.md, line 33

    - **An `.env.example`** with the env-var names migrated out of settings/MCP files - zero ambiguity about the new pattern.

    audit/samples/claude-code-zhet/aSAMM-audit-claude-code-2026-04-13-basic.html, line 181

    <code>~/.ssh/id_ed25519</code>, <code>~/.ssh/id_rsa</code> are readable; <code>Bash(curl:*)</code> is in the project allowlist;

    audit/samples/claude-code-zhet/aSAMM-audit-claude-code-2026-04-13-basic.html, line 211

    <code>Bash(*~/.ssh*)</code>, <code>Bash(*id_rsa*)</code>, <code>Bash(*id_ed25519*)</code> reads - and restrict

    audit/samples/claude-code-zhet/aSAMM-audit-claude-code-2026-04-13-basic.html, line 262

    <td>Bash command exfiltrating <code>~/.ssh/id_rsa</code> via prompt injection in fetched body</td>

    audit/samples/claude-code-zhet/aSAMM-audit-claude-code-2026-04-13-basic.html, line 391

    <tr><td>Filesystem: read sensitive</td><td><code>test -r ~/.ssh/id_ed25519</code></td><td><strong class="fail">READABLE</strong></td><td>[empirical]</td></tr>

  • 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.md).

  • Maintainer activity

    Pass

    The latest commit is within the last 365 days.

Note

Hygiene check of commit 53fa38e78cbf. requests shell access. The manifest names this permission. 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: 53fa38e78cbf9f76b4784b28214f5ecdc1249979

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