Caveat verdict

mjolnir-brain

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

Local-first AI agent memory system using Markdown and JSON files; network backup features are explicitly opt-in and disabled by default, requiring manual credential configuration via env vars.

⚠ 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
100
transparency
80
maintenance

Findings (9)

Pattern match critical

Pipe to python — executes piped content as Python code

INSTALL.md · code · | python3

Pattern match critical

Recursive delete from root or home — destructive command

INSTALL.md · code · rm -rf ~

Pattern match high

Accesses .ssh directory

docs/工作记录机制.md · code · .ssh/

Pattern match medium

References agent memory files

docs/architecture.md · code · MEMORY.md

Pattern match medium

Sets world-executable permissions

docs/multi-user.md · code · chmod 700

Pattern match medium

Opens WebSocket connection

docs/v3.0-architecture.md · code · WebSocket

Pattern match medium

References sudo — requests elevated privileges

docs/security.md · prose · downgraded · sudo

Pattern match medium

Accesses shell history/config

playbooks/multi-user-setup.md · prose · downgraded · ~/.zshrc

Pattern match medium

Accesses OpenClaw config/secrets directly

templates/TOOLS.md · prose · downgraded · ~/.openclaw/openclaw.json

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class B/F). 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.

Is this flag fair?

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