SKILLEMALL.ai
GitHub ↗ RU Sign in

BC skill-injection-defense

Audits legal AI skills, prompts, workflows, MCP/tool instructions, and agent packages for prompt injection, malicious instructions, unsafe scripts, suspicious metadata, credential exposure, exfiltration paths, persistence mechanisms, and supply-chain risk. Use before installing, importing, publishing, promoting, modifying, or trusting third-party or generated AI skills and legal workflows. Do not use as a generic code review unless skill trust, workflow safety, prompt injection, or supply-chain hygiene is in scope.

lawve-ai/awesome-legal-skills Agent Skills author: lawve-ai 2 files body ≈ 986 tokens Open the sourcegithub.com↗ analyzed 75 min ago

Audits legal AI skills, prompts, workflows, MCP/tool instructions, and agent packages for prompt injection, malicious instructions, unsafe scripts, suspicious…

As a process C 57/100 · Has gaps — weak spots: inputs and preconditions, failures and branches, consistency

ProcedureAI and agentstype and topics are labelled automatically from the skill text
Runs in: 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) OpenClaw Goose OpenCode
JSON
Technical rating
B
94/100
safety, quality, tests
Safety 60%
98
Quality 40%
87
Run on models
none yet
Process rating
C
57/100
Has gaps
Inputs and preconditions w 11
0
Failures and branches w 10
0
Progress reporting w 2
0
the three weakest of ten parameters · all ten

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 · 2

    ✓ No critical or high findings

    Medium and low: 2
    • low Risky intent intent-offensive-security SKILL.md:43
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      - credential harvesting or environment-variable access;
    • low Risky intent intent-offensive-security SKILL.md:46
      Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
      - destructive actions such as deletion, overwrite, privilege escalation, or broad filesystem mutation;

    Files scanned: 2. 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 57/100

    • 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
    • 30Running it twice. 6 mutating operations with no state check
    • 40Consistency. Frontmatter name (skill-injection-defense) differs from the folder (skill-injection-defense-adrian-lerer)
    • 60Tools and files. Uses tools (bash) that frontmatter does not declare
    • 60Result and completion. Output format stated, no completion criterion
    • 100Steps. 69 steps
    • 100When it triggers. States when to use and when not to
    • 100Execution cost. Instruction body is 986 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
    • +4No input/output examples
    • +1No license
    • +2Single-language instructions
    • +4Description says when NOT to use the skill
    • +3Description length 520: enough signal without eating the budget
    • +4Structure: 7 headings
    • +3Step-by-step instructions: 69 items
    • +3Output format is stated explicitly

    Quality base 70; lint remarks subtract, signals add up to 100. Result: 87.