wiki-update

Pass

Audited by Gen Agent Trust Hub on Nov 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the current working directory (README files, source code, documentation, and git logs) to distill architectural insights. This information is then used to populate wiki pages and passed as arguments to shell commands for synchronization logging. A maliciously crafted project file could influence the distillation process or metadata generation.
  • Ingestion points: README.md, docs/ folder, source code files, package.json, pyproject.toml, and git commit history are scanned in the Understand the Project step.
  • Boundary markers: Absent. The instructions do not define delimiters or provide specific 'ignore' directives for the content being distilled from the project.
  • Capability inventory: The skill has the ability to write to the local file system (Obsidian vault), execute shell commands (git, obsidian-wiki, qmd), and optionally trigger package installations (npm).
  • Sanitization: Absent. Content extracted from the project is interpolated into wiki templates and CLI arguments without explicit sanitization or escaping of control characters.
  • [EXTERNAL_DOWNLOADS]: The skill suggests the installation of an external Node.js package @colbymchenry/codegraph via npm install -g. While this is presented as an optional optimization and requires user consent, it introduces a dependency on an external, third-party package.
  • [COMMAND_EXECUTION]: The skill relies on executing several shell commands to gather project metadata and sync logs, including git (for retrieving remote URLs and commit history), obsidian-wiki (for code understanding and memory synchronization), and qmd (for search indexing).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 10, 2026, 05:22 PM