wiki-connect

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it ingests and processes untrusted wiki content.
  • Ingestion points: Wiki state is loaded via maestro wiki commands in Stage 1, which brings external text into the agent's context.
  • Boundary markers: No explicit delimiters (e.g., XML tags or specific markers) or 'ignore embedded instructions' warnings are provided to separate wiki data from the skill's instructions.
  • Capability inventory: The skill possesses the ability to execute shell commands (maestro wiki update via the Bash tool) and perform file writes/edits in Stage 5 and 6.
  • Sanitization: There is no evidence of sanitization, escaping, or validation of the ingested wiki content before it is used for scoring or displayed to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 08:44 AM
Security Audit — agent-trust-hub — wiki-connect