ClawAudit verdict
OpenCortex
opencortex
Self-improving memory architecture with nightly distillation; sensitive features like voice profiling, infrastructure collection, and git push are explicitly OFF by default and require opt-in via environment variable.
⚠ 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.
Permission integrity
agent_memory
Findings (5)
Possible hardcoded credential
scripts/git-backup.sh · prose · downgraded · secret="$(printf
Pipe to python — executes piped content as Python code
scripts/verify.sh · prose · downgraded · | python3
References agent memory files
SKILL.md · code · MEMORY.md
Accesses system credential store
README.md · prose · downgraded · Keychain
Sets world-executable permissions
scripts/vault.sh · prose · downgraded · chmod 700
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
Requires 3 system binaries.
agent_memory Is this flag fair?
Thanks — recorded.