architect-fast

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill is configured to execute local management scripts (status.ps1 and status.sh) located within the project's internal directory structure for status tracking.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted external information, such as raw test suite results and user-defined issue requirements, which are subsequently passed to subagents (builders) to guide their work.
  • Ingestion points: Raw output from the full test suite and issue bodies containing acceptance criteria (referenced in SKILL.md Procedure steps 5 and 6).
  • Boundary markers: Employs specific calibration instructions to restrict the subagent's review focus, such as "Flag only gaps that affect correctness... Do not report stylistic preferences."
  • Capability inventory: Orchestrates subagent lifecycles, performs branch merges, and executes local status scripts.
  • Sanitization: No explicit sanitization or filtering is described for the raw test output before it is processed by the subagents.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 10:07 AM