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

ruminate

Pass

Audited by Gen Agent Trust Hub on Mar 22, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill accesses sensitive user data by reading conversation logs from ~/.claude/projects/ and ~/.codex/sessions/. This exposure of private chat history is necessary for the skill's primary function but involves handling highly personal information.
  • [COMMAND_EXECUTION]: The process involves executing a shell script from another skill (.claude/skills/meditate/scripts/snapshot.sh) and a local Python script for data extraction. These executions are confined to the local environment.
  • [PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks when processing untrusted historical data.
  • Ingestion points: The skill ingests raw text from past conversation JSONL files via the extract-conversations.py script.
  • Boundary markers: No explicit delimiters or 'ignore' instructions are used in the prompts for the analysis sub-agents to prevent the execution of instructions found within the logs.
  • Capability inventory: The skill possesses the ability to spawn sub-agent teams (TeamCreate) and execute local scripts.
  • Sanitization: No sanitization or filtering of the historical message content is performed prior to analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 22, 2026, 01:17 AM