bet-invalidation-checkpoint
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill functions as a localized document processor for business performance reviews. It restricts file interactions to relative project directories such as context/ and reviews/. No patterns of command execution, privilege escalation, or persistence were detected.
- [PROMPT_INJECTION]: The skill exhibits a standard surface for indirect prompt injection due to its document generation logic. 1. Ingestion points: The agent ingests metric values and evidence provided by the user or from local bet records. 2. Boundary markers: The skill uses structured Markdown templates but does not employ specific delimiters or instruction-guards for user-supplied strings. 3. Capability inventory: Access is limited to reading and writing Markdown files and triggering internal task-based skills like /decision-record. 4. Sanitization: No explicit content filtering or escaping is defined in the instructions. This surface is inherent to the skill's purpose and presents minimal risk given the lack of dangerous capabilities.
Audit Metadata