team-release

Pass

Audited by Gen Agent Trust Hub on Sep 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project state and milestone files to automatically determine the release version, which represents a potential ingestion surface for instructions embedded in external files.
  • Ingestion points: production/session-state/active.md and the most recent files in the production/milestones/ directory.
  • Boundary markers: The skill implements a strict confirmation protocol, using AskUserQuestion to verify inferred version numbers with the user before proceeding.
  • Capability inventory: The skill is authorized to use Bash, Write, Edit, and Task tools, the latter of which is used to spawn specialized sub-agents.
  • Sanitization: No automated content filtering is described; the skill relies on the user to review and approve the context before execution.
  • [SAFE]: The orchestrator incorporates strong security practices by enforcing an error recovery protocol that surfaces blocks immediately to the user and requires manual approval for overriding a 'NO-GO' decision from the release sub-agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 29, 2026, 04:31 PM