today

Pass

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands grep and git log to discover tasks and activity within the local project environment.
  • [PROMPT_INJECTION]: The skill ingests content from files that could contain malicious instructions designed to influence the agent's behavior (indirect prompt injection).
  • Ingestion points: Content is read from tasks/*.md, projects/*.md, and existing daily/ files.
  • Boundary markers: No markers are present to delimit untrusted content from the rest of the prompt.
  • Capability inventory: The agent has access to read and write files and execute shell commands (grep, git log) as part of its planning operations.
  • Sanitization: No sanitization or validation of the ingested content is performed before it is processed by the agent.
  • Mitigation: Use delimiters like XML tags or labeled sections to separate vault content and include instructions to the agent to ignore any embedded directives within those sections.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 20, 2026, 06:48 AM
Security Audit — agent-trust-hub — today