trim-spec

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external documents such as specifications and Architecture Decision Records (ADRs), which act as untrusted data ingestion points.\n
  • Ingestion points: Reads content from file paths or folders provided as the 'Target' in the input (SKILL.md).\n
  • Boundary markers: The instructions do not define explicit boundary markers or delimiters when passing raw file content to the general-purpose agents for extraction in Step 1 or loss verification in Step 4.\n
  • Capability inventory: The skill performs file reads and in-place file writes, and invokes other agent skills including reviewer, general-purpose, and tighten-instruction (SKILL.md).\n
  • Sanitization: No content sanitization or validation is specified to filter out potential instructions embedded within the processed text documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 10:35 AM
Security Audit — agent-trust-hub — trim-spec