spec-gathering

Pass

Audited by Gen Agent Trust Hub on Nov 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local shell commands in SKILL.md and tests/spec-gathering-integration.test.md to read project context and verify the creation of specification files. These operations are limited to standard file system checks and grep commands within the local environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied task descriptions, creating a surface for indirect prompt injection. This risk is addressed through a multi-phase workflow that includes:\n
  • Ingestion points: User input is captured in Phase 2 and Phase 5 of SKILL.md.\n
  • Boundary markers: Explicit confirmation is required in Phase 6 before any artifacts are generated.\n
  • Capability inventory: The skill uses Bash for file inspection and template-renderer for file writing.\n
  • Sanitization: Relies on human-in-the-loop verification to validate requirements and assumptions.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 6, 2026, 03:31 PM