ClawAudit verdict
multi-agent-memory
light-office-multi-agent-memory
The skill appears to be a multi-agent memory system with various features such as auto-capture, RRF retrieval, knowledge graph, and token tracking. Although it has a wide range of capabilities, they seem to be used for legitimate purposes.
⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
Findings (4)
subprocess execution — runs system commands from Python
scripts/memory-system.py · prose · downgraded · subprocess.run(
Contains shortened/invite URL
SKILL.md · code · discord.gg/light
Python os.environ.get — reads environment variable
scripts/agent-manager.py · prose · downgraded · os.environ.get(
References agent configuration files
scripts/agent-manager.py · prose · downgraded · AGENT_CONFIG
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
Requires 2 system binaries. (1 elevated: git).
Is this flag fair?
Thanks — recorded.