factory-intent

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The orchestrator invokes local Bash and PowerShell scripts to automate the software development lifecycle. These scripts, sourced from the factory-mission skill, are located in the .agents/skills/factory-mission/scripts/ directory. Specific commands include run-branch.sh for environment setup, commit-step.sh for versioning artifacts, and open-mr.sh for publishing changes to the remote repository.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted 'Mission Brief' documents to generate complex project documentation (PRD and AD files).
  • Ingestion points: The Mission Brief enters the agent's context as the source material for the specify phase in both product and architecture lanes.
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the brief's contents as data only and to ignore any embedded instructions.
  • Capability inventory: The skill possesses significant write capabilities to project directories and the ability to perform Git operations, including pushing to the remote origin.
  • Sanitization: No input validation or content filtering is implemented for the mission brief, creating a risk that instructions embedded in the document could influence the agent's behavior during the implementation and analysis phases.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 03:59 PM