spec-diff-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of technical specifications, versioned files, and git diffs.
- Ingestion points: The agent reads user-provided paths for specifications and supporting documents within the repository root (e.g., spec_path, version_a_path).
- Boundary markers: There are no explicit instructions or delimiters to isolate untrusted specification content from the agent's operational logic.
- Capability inventory: The skill writes planning artifacts to the filesystem (design/changes/) and generates implementation tasks for downstream agents like sphere-feature-workflow, creating a vulnerability surface where malicious spec content could influence subsequent development steps.
- Sanitization: No sanitization, validation, or filtering of external content is specified before processing.
Audit Metadata