topic-synthesis
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
monologueCLI tool (monologue notes all,monologue notes get) to search for and retrieve the content of user recordings. These operations are restricted to read-only discovery and retrieval within the local environment. - [CREDENTIALS_UNSAFE]: The instructions include a proactive security measure that forbids the agent from requesting or accepting API tokens within the chat interface, instead pointing the user toward a local onboarding process.
- [DATA_EXFILTRATION]: The skill contains explicit guardrails against the unauthorized transmission of data. It restricts the agent to read-only access via the
$monologue-notesinterface and prohibits the creation of any external records, messages, or tasks without the user's explicit approval.
Audit Metadata