agent-wiki

Warn

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands including git init, git submodule add, and git push to manage the documentation lifecycle, and runs a local script with node for PDF processing.
  • [CREDENTIALS_UNSAFE]: Documentation in workflow/update-mode.md and evals/evals.json shows the skill attempts to read Git authentication credentials from the local file path ~/.config/agent-wiki/credentials.
  • [EXTERNAL_DOWNLOADS]: The skill triggers the installation of the pdf-parse library from the official npm registry to enable PDF text extraction functionality.
  • [PROMPT_INJECTION]: The skill processes unstructured data from external sources such as RFPs and meeting notes, which creates a surface for indirect prompt injection where malicious instructions could be embedded in the source documents.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 10, 2026, 02:20 AM