LLM Wiki Maintainer
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides process guidelines for maintaining a local wiki, including reading indices, logging updates, and organizing project-specific knowledge.
- [COMMAND_EXECUTION]: The skill references 'LLM Wiki plugin tools' for operations like source reads, page writes, and patch proposals. These tools are part of the intended environment and no arbitrary command execution was found.
- [DATA_EXPOSURE]: The skill instructions involve reading local files such as 'AGENTS.md', 'wiki/index.md', and 'wiki/log.md'. This access is limited to the local wiki scope and is necessary for the skill's primary function of wiki maintenance.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external source material and wiki content. While this creates a surface for indirect prompt injection if the source data contains malicious instructions, the skill does not exhibit dangerous autonomous capabilities that would escalate this risk beyond a low level.
Audit Metadata