wiki-mode

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes local Python helper scripts (claude-obsidian.py and wiki-mode.py) located within the application's installation directory to perform routing logic and configuration updates. These executions are constrained to a fixed product root and specific command-line arguments.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from a user-controlled configuration file (.vault-meta/mode.json) to determine the active methodology. Ingestion points: Configuration data is read from the .vault-meta/mode.json file. Boundary markers: The skill uses strict validation of mode names but lacks explicit delimiters for the configuration file contents. Capability inventory: The skill can execute local scripts and read/write vault metadata. Sanitization: The skill implements a SHA256-based plan verification mechanism, ensuring changes are reviewed and consistent with the intended state.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 04:50 PM
Security Audit — agent-trust-hub — wiki-mode