Caveat verdict

mind-wander

88
🟢 Trusted
No high-risk patterns surfaced by the deep scan — automated capability review, not behavioral proof.

Background autonomous research agent that uses a local LLM (Qwen3.5-9B via Ollama), reads a local markdown file for questions, writes findings to local files, and optionally queries Perplexity for web search — all behavior matches the stated purpose with no exfiltration or injection.

⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.

Automated static analysis — not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.

0
security
50
transparency
70
maintenance

Permission integrity

Performs file operations but does not declare file-accessing binaries

file_read+write

Findings (10)

Pattern match critical

Possible hardcoded credential

references/setup.md · code · TOKEN = "your-gateway-token

Pattern match high

HTTP request to bare IP address — common in malicious payloads

SKILL.md · code · http://172.18.0.1

Pattern match high

subprocess execution — runs system commands from Python

references/setup.md · code · subprocess.run(

Pattern match high

Uses eval() — can execute arbitrary code

scripts/tools.py · prose · downgraded · eval(

Pattern match medium

Python httpx request — network access

references/setup.md · code · httpx.post(

Pattern match medium

Accesses OpenClaw config/secrets directly

scripts/install.sh · prose · downgraded · ~/.openclaw/openclaw.json

Pattern match medium

Uses exec() — may execute shell commands

scripts/tools.py · prose · downgraded · exec(

Pattern match low

pip3 install — installs Python packages at runtime

scripts/install.sh · prose · downgraded · pip3 install

Pattern match low

Python os.environ.get — reads environment variable

scripts/mind_wander_config.py · prose · downgraded · os.environ.get(

Pattern match low

POSTs data to external URL

scripts/tools.py · prose · downgraded · .post( "https://

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class D). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.

Permissions & capabilities

No declared permissions — minimal attack surface.

file_write

Is this flag fair?

Check another skill Browse the registry Auditing your own skills or configs? Use the API