ecosystem-managing

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (LOW): The skill specifies the use of npm run build and grep_search to verify system state and find references during the 'PURGE' protocol. While standard for developers, these involve executing shell commands that could be exploited if inputs are manipulated.- [INDIRECT PROMPT INJECTION] (LOW): The script version_control.py provides a surface for indirect injection. Evidence Chain: 1. Ingestion points: The changes_summary parameter in update_registry_with_changelog (scripts/version_control.py). 2. Boundary markers: Absent; the summary is directly appended to the file. 3. Capability inventory: File write access (open(..., 'w')) to project files. 4. Sanitization: Absent; the input string is concatenated directly into the Markdown content.- [PRIVILEGE ESCALATION] (SAFE): The 'PURGE' protocol describes deleting code, documents, and tests. While destructive, this capability is aligned with the skill's primary stated purpose of ecosystem management and does not involve unauthorized privilege gains such as sudo usage.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:44 PM