1k-coding-patterns

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface detected.
  • Ingestion points: The skill reads external code using Read, Grep, and Glob tools.
  • Boundary markers: No delimiters or instructions are provided to the agent to ignore instructions embedded within the files it processes.
  • Capability inventory: The skill has Write and Edit tools, which allow it to modify files on the system.
  • Sanitization: No sanitization or validation of input data is present.
  • [EXTERNAL_DOWNLOADS]: The skill is installed from an untrusted third-party repository (majiayu000/claude-skill-registry). The install command downloads the main skill file, but metadata suggests the presence of scripts not included in the provided analysis files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 09:54 PM