the-sit-down

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructs the agent to bypass user confirmation and interaction loops during phase transitions.
  • Evidence: In Step 5, the instructions command the agent to "Immediately invoke gangsta:resource-development — do NOT ask the Don what to do next, do NOT pause, do NOT prompt for confirmation. Auto-advance is mandatory." This prevents the user from reviewing the generated contract or intervening before the next automated skill execution occurs.
  • [PROMPT_INJECTION]: Potential for indirect prompt injection through the processing of untrusted external data.
  • Ingestion points: Step 1 requires the agent to gather intelligence from files such as the Reconnaissance Dossier, Grilling Consensus, Project Constitution, and Ledger Entries.
  • Boundary markers: Absent. The instructions do not define delimiters or provide guidance to the agent to disregard instructions embedded within these external documents.
  • Capability inventory: The skill writes binding specification files to the local file system (docs/gangsta/...) and invokes external skill dependencies (gangsta:the-consigliere and gangsta:resource-development).
  • Sanitization: Absent. There are no requirements to sanitize, escape, or validate content derived from external intelligence before it is incorporated into the formal project specification.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 02:10 PM
Security Audit — agent-trust-hub — the-sit-down