digikam-search

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a specific local binary at /Users/mtm/go/bin/islandiguana. This tool is used to query an Obsidian vault. The use of absolute paths for binaries and data directories is specific to a particular user environment and can be a security concern if the environment is shared or the binary integrity is unverified.
  • [PROMPT_INJECTION]: The skill processes content from an external source (Obsidian vault notes), making it susceptible to indirect prompt injection if those notes contain malicious instructions.
  • Ingestion points: Local files located within the /Users/mtm/Documents/Obsidian Vault directory.
  • Boundary markers: None. The skill directly executes a command to retrieve and then 'read the returned notes'.
  • Capability inventory: Shell command execution via the islandiguana binary.
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the retrieved note content before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 12:40 AM
Security Audit — agent-trust-hub — digikam-search