skills/ymd38/dev-skills/spec-doc/Gen Agent Trust Hub

spec-doc

Pass

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • Prompt Injection (LOW): The skill is vulnerable to Indirect Prompt Injection because it ingests untrusted data from the local file system without adequate sanitization or boundary markers.
  • Ingestion points: The agent performs a comprehensive scan of the codebase, including entry points, data models, public interfaces, and configuration files (Phase 1.2: Code Scan Strategy).
  • Boundary markers: The instructions lack specific delimiters or warnings to ignore instructions embedded within the source code (e.g., in code comments or string literals).
  • Capability inventory: The skill has broad file-read access to the directory and file-write access to the docs/ folder.
  • Sanitization: There is no logic to escape or validate content extracted from the code before it is interpolated into the final markdown specification.
  • Data Exposure (SAFE): While the skill scans configuration files and environment variable definitions, it does so for documentation purposes. There are no network operations detected to exfiltrate this data.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 19, 2026, 05:19 PM