SKILLEMALL.ai
GitHub ↗ RU Sign in

BD crabukit

Security scanner for OpenClaw skills with Clawdex integration. Analyzes SKILL.md and scripts for dangerous permissions, hardcoded secrets, shell injection vulnerabilities, and malicious code patterns. Automatically uses Clawdex database if installed for known-malicious skill detection. Use when (1) installing a skill from an untrusted source, (2) developing a skill before publishing, (3) auditing installed skills, or (4) running CI/CD security checks.

modbender/skill-library-mcp Agent Skills author: modbender MIT 29 files · 1 script body ≈ 617 tokens Open the sourcegithub.com↗ analyzed 2 d ago

Security scanner for OpenClaw skills with Clawdex integration.

As a process D 46/100 · Unfinished process — weak spots: result and completion, when it triggers, inputs and preconditions

IntegrationSoftware developmentInfrastructuretype and topics are labelled automatically from the skill text
Runs in: OpenClaw Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) Goose OpenCode
JSON
Technical rating
B
80/100
safety, quality, tests
Safety 60%
79
Quality 40%
81
Run on models
none yet
Process rating
D
46/100
Unfinished process
Result and completion w 14
0
Inputs and preconditions w 11
0
Failures and branches w 10
0
the three weakest of ten parameters · all ten

What is at stake

Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.

Instruction override medium severity

Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.

If you install

The text contains phrases like "ignore previous instructions" or "you are now…". That is an attempt to hijack the agent: it may break your rules, the system limits or company policy.

For the author

An honest skill does not need them: state the role and the rules directly without overriding other instructions. Otherwise catalog scanners and corporate filters will block the listing.

How to improve

    For the model run — optional
    • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
    • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

    Guard findings · 17

    ✓ No critical or high findings

    Medium and low: 17
    • medium Instruction override en-ignore-previous crabukit/rules/patterns.py:389
      Instruction-override phrase ("ignore previous instructions") (detector / deny-list definition)
      "description": "Contains pattern attempting to override previous instructions",
      detector
    • low Dangerous commands cmd-background-process crabukit/analyzers/bash_static.py:395
      Starts a background / autostarted process (string literal in code, not executed)
      if '/etc/systemd/system' in line or 'systemctl enable' in line:
      code literal
    • low Dangerous commands cmd-privilege crabukit/analyzers/bash_static.py:420
      Privilege escalation / world-writable permissions (code comment)
      # Check for setuid/setgid
      comment
    • low Dangerous commands cmd-privilege crabukit/analyzers/bash_static.py:424
      Privilege escalation / world-writable permissions (detector / deny-list definition; string literal in code, not executed)
      title="setuid/setgid bit set",
      detectorcode literal
    • low Dangerous commands cmd-privilege crabukit/analyzers/bash_static.py:425
      Privilege escalation / world-writable permissions (string literal in code, not executed)
      description="Setting setuid/setgid bits can create privilege escalation vulnerabilities",
      code literal
    • low Dangerous commands cmd-privilege crabukit/analyzers/bash_static.py:430
      Privilege escalation / world-writable permissions (string literal in code, not executed)
      remediation="Avoid setuid/setgid; use proper privilege separation",
      code literal
    • low Dangerous commands cmd-pipe-to-shell crabukit/rules/patterns.py:244
      Downloads and executes remote code from an unrecognised host (pipe to shell) (detector / deny-list definition; string literal in code, not executed)
      "title": "curl | bash pattern",
      detectorcode literal
    • low Dangerous commands cmd-privilege crabukit/rules/patterns.py:319
      Privilege escalation / world-writable permissions (string literal in code, not executed)
      "description": "chmod 777 grants full read/write/execute to all users",
      code literal
    • low Dangerous commands cmd-shell-rc README.md:138
      Writes to a shell startup file (documentation of a security skill)
      echo "source ~/.claw-safe-install.sh" >> ~/.zshrc
      security skill
    • low Instruction override en-ignore-previous RESEARCH_SUMMARY.md:17
      Instruction-override phrase ("ignore previous instructions") (documentation table row; documentation of a security skill)
      | `ignore_instructions` | "Ignore all previous instructions" | HIGH |
      tablesecurity skill
    • low Risky intent intent-offensive-security RESEARCH_SUMMARY.md:18
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      | `developer_mode` | Privilege escalation attempts | HIGH |
    • low Dangerous commands cmd-privilege RESEARCH_SUMMARY.md:95
      Privilege escalation / world-writable permissions (documentation of a security skill)
      - **Privilege escalation**: setuid/setgid, sudo patterns
      security skill
    • low Risky intent intent-offensive-security RESEARCH_SUMMARY.md:95
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      - **Privilege escalation**: setuid/setgid, sudo patterns
    • low Dangerous commands cmd-shell-rc scripts/README.md:18
      Writes to a shell startup file (documentation of a security skill)
      echo "source ~/.claw-safe-install.sh" >> ~/.zshrc
      security skill

    A further 3 matches are quotations in this security skill's documentation and are not counted as findings.

    Files scanned: 26. Evidence is masked. Grey chips explain why severity was lowered.

    Against the Agent Skills spec

    ✓ No remarks against the Agent Skills spec

    Process rating: all ten parameters 46/100

    • 0Result and completion. Does not say what the result is
    • 0Inputs and preconditions. Does not say what the process needs to start
    • 0Failures and branches. Linear process with no failure handling
    • 0Progress reporting. Says nothing while it works
    • 20When it triggers. No condition that starts the skill
    • 60Tools and files. Uses tools (bash, web) that frontmatter does not declare
    • 100Steps. 4 steps
    • 100Consistency. Name and required fields are in place
    • 100Execution cost. Instruction body is 617 tokens
    • 100Running it twice. No mutating operations

    Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

    Quality signals

    • +5Description has no quoted example phrases that should trigger the skill
    • +4Description does not say when NOT to use the skill (false activations)
    • +3Output format is not stated: the model decides each time
    • -31 of 1 scripts are never mentioned in SKILL.md
    • +1No license
    • +2Single-language instructions
    • +3Description length 455: enough signal without eating the budget
    • +4Structure: 8 headings
    • +3Step-by-step instructions: 4 items
    • +4Has examples (4 code blocks)

    Quality base 70; lint remarks subtract, signals add up to 100. Result: 81.