skills/arcblock/idd/intent-critique/Gen Agent Trust Hub

intent-critique

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_CONTEXT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Uses local shell utilities wc -l and git log to calculate line counts and check commit history for design analysis metrics.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes content from user-provided design files such as INTENT.md and records/INDEX.md. As a design critique tool, this ingestion is a core functional requirement. The analysis focuses on structural patterns (complexity, YAGNI, over-engineering) rather than executing embedded instructions.\n- [DYNAMIC_CONTEXT_INJECTION]: Employs shell command placeholders (e.g., !wc -l INTENT.md!) to dynamically calculate document statistics at load time. These commands are limited to local metadata retrieval for the purpose of design auditing.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 12:42 PM