SKILLEMALL.ai
GitHub ↗ RU Sign in

BD skill-auditor

Security scanner for OpenClaw skills. Detects malicious code, obfuscated payloads, prompt injection, social engineering, typosquatting, and data exfiltration before installation. Features 0-100 numeric risk scoring, MITRE ATT&CK mappings, base64/hex deobfuscation, IoC database, whitelist system, and SHA256 file inventory. Use before installing any third-party skill. Triggers: audit skill, check security, scan skill, is this skill safe, security review, quarantine.

modbender/skill-library-mcp Agent Skills author: modbender MIT 6 files · 1 script body ≈ 862 tokens Open the sourcegithub.com↗ analyzed 3 d ago

Security scanner for OpenClaw skills.

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

AnalyzerAI and agentsSoftware developmenttype 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
86/100
safety, quality, tests
Safety 60%
82
Quality 40%
92
Run on models
none yet
Process rating
D
46/100
Unfinished process
Result and completion w 14
0
Failures and branches w 10
0
Progress reporting w 2
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.

Exfiltration 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 instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".

For the author

If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.

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 · 14

    ✓ No critical or high findings

    Medium and low: 14
    • medium Exfiltration intent-browser-credential-store references/ioc-database.json:108
      Accesses a browser credential / cookie store (detector / deny-list definition)
      "pattern": "Cookies|Login Data|Web Data|chrome/Default|firefox/profiles",
      detector
    • low Risky intent intent-offensive-security references/known-patterns.md:3
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      ## C2 Server IPs
    • low Risky intent intent-offensive-security references/known-patterns.md:19
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      | burpcollaborator.net | Pentest exfil | MEDIUM |
    • low Risky intent intent-wallet-secrets scripts/audit_skill.py:215
      Handles crypto-wallet secrets (seed / mnemonic / private key) — a classic stealer target (detector / deny-list definition)
      (r'(?:seed|mnemonic)\s*(?:phrase|words?)', "seed/mnemonic phrase reference"),
      detector

    A further 10 matches are quotations in this security skill's documentation and are not counted as findings.

    Files scanned: 6. 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
    • 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
    • 30Running it twice. 1 mutating operations with no state check
    • 40Consistency. Frontmatter name (skill-auditor) differs from the folder (skill-auditor-v2)
    • 60Tools and files. Uses tools (bash, python) that frontmatter does not declare
    • 70Inputs and preconditions. Inputs and preconditions are listed
    • 100Steps. 23 steps
    • 100Execution cost. Instruction body is 862 tokens

    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
    • +2Single-language instructions
    • +3Description length 468: enough signal without eating the budget
    • +4Structure: 16 headings
    • +3Step-by-step instructions: 23 items
    • +4Has examples (4 code blocks)
    • +4Reference files are cited in the instructions (2 of 2)
    • +3All 2 scripts are documented
    • +1License stated

    Quality base 70; lint remarks subtract, signals add up to 100. Result: 92.