wiki-architect

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by design as it ingests and processes untrusted data from repository file structures and contents.
  • Ingestion points: The skill scans repository file trees, README files, and source code contents to generate documentation structures (SKILL.md).
  • Boundary markers: No delimiters or instructions to ignore embedded commands are specified for the processed content, which may lead the agent to follow malicious instructions hidden in a repository.
  • Capability inventory: Analysis of all procedures shows the skill is restricted to reading file structures and outputting JSON documentation; it has no capabilities for network operations, file writing, or command execution.
  • Sanitization: There are no documented mechanisms for sanitizing or validating the data retrieved from external codebases before it is used in documentation prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 06:51 AM