wiki-manager
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes content from GitHub Wiki pages which are external, potentially untrusted sources. \n
- Ingestion points: Content is read from markdown files in the cloned wiki repository during 'Read Page', 'Edit Page', and 'Search' operations. \n
- Boundary markers: The skill instructions lack explicit boundary markers or instructions for the agent to ignore potentially malicious content embedded in wiki pages. \n
- Capability inventory: The agent can perform file writes, deletes, and git operations (commit, push), which could be misused if influenced by malicious content. \n
- Sanitization: There is no mention of sanitizing or validating the content of the wiki pages before the agent processes or outputs them.\n- [COMMAND_EXECUTION]: The skill utilizes system CLI tools to manage wiki content and authentication. \n
- Evidence: The workflow involves executing
gh api userfor identity verification and variousgitcommands (clone, commit, push) to manage the wiki repository.
Audit Metadata