update-knowledge-base

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill ingests data from external sources that could contain instructions designed to manipulate the agent's behavior during the documentation process.
      1. Ingestion points: Reads release notes via 'gh release list' and project data from Linear and Jira tools.
      1. Boundary markers: Absent. The execution strategy does not specify delimiters or instructions to ignore embedded prompts within the retrieved data.
      1. Capability inventory: The skill can write files to the local filesystem (05-knowledge/) and sync updates to external wikis like Notion and Confluence.
      1. Sanitization: Absent. Content gathered from external sources is directly used to populate templates and update documentation without validation.
  • [COMMAND_EXECUTION]: Execution of system commands and CLI tools.
    • Evidence: Runs 'date' and 'mkdir -p' using Bash for environment management.
    • Evidence: Uses 'gh release list' to fetch external repository data.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 11:30 AM