llm-wiki-github-action
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill involves reading and validating files within the repository (such as skills frontmatter and documentation), which creates a potential surface for instructions embedded in those files to influence the agent's behavior during maintenance tasks.
- Ingestion points: The skill reads
skills/*/SKILL.mdand other repository documents during validation, linting, and report generation procedures (SKILL.md). - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified for the ingestion process.
- Capability inventory: The skill facilitates file-system modifications through branch creation, committing generated reports/docs, and opening pull requests (SKILL.md).
- Sanitization: The skill includes specific "Safety gates" and templates for "security scans" and "redaction/security checks" to mitigate risks associated with automated processing of repository content. The risk is assessed as safe given the maintenance context and the requirement for human review of all automated output.
Audit Metadata