Caveat verdict
proactive-agent
A behavioral framework for proactive agent architecture with only agent_memory capability; no external data exfiltration, no execution sinks, and no manipulation attempts — all guidance is transparent and user-serving.
⚠ 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.
Permission integrity
agent_memory
Findings (5)
Prompt injection — tries to override agent instructions
references/security-patterns.md · code · Ignore previous instructions
Raw model control tokens — prompt injection via token manipulation
references/security-patterns.md · code · [INST]
Possible prompt injection — attempts to redefine agent identity
references/security-patterns.md · code · You are now
References agent memory files
SKILL.md · code · MEMORY.md
Popular HTTP library — network access
SKILL.md · code · Got
Permissions & capabilities
No declared permissions — minimal attack surface.
agent_memory Is this flag fair?
Thanks — recorded.