skills/poteto/noodle/meditate/Gen Agent Trust Hub

meditate

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its automated processing of project content.
  • Ingestion points: Markdown files from the 'brain/', '.agents/skills/', and '~/.claude/projects//memory/' directories are aggregated into a single snapshot (Step 1 in SKILL.md).
  • Boundary markers: Files are separated by '=== path ===' headers in the snapshot script, but no explicit instructions are provided to sub-agents to ignore instructions contained within the ingested data.
  • Capability inventory: The skill can autonomously apply changes, including updating or deleting files based on sub-agent reports (Step 6 in SKILL.md).
  • Sanitization: The aggregated content is not sanitized or validated before being analyzed by sub-agents.
  • [COMMAND_EXECUTION]: The skill executes a local shell script 'scripts/snapshot.sh' to generate snapshots of project data. The script uses standard utilities like find and cat to collect file contents and does not interact with the network.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 01:35 AM