lode-session-start-recall

Pass

Audited by Gen Agent Trust Hub on May 12, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a bundled Python script to process and retrieve project data.
  • Evidence: SKILL.md contains the instruction to run python <this-skill>/scripts/recall_context.py --cwd "$PWD" --limit 12.
  • [SAFE]: The skill accesses local configuration and history files within the Lode system's designated directories (~/.lode, ~/.weekly-ppt). No sensitive system credentials or unauthorized data exfiltration patterns were detected.
  • [SAFE]: The skill ingests data from local JSON files containing project history, which acts as an indirect prompt injection surface.
  • Ingestion points: scripts/recall_context.py reads data from {vault}/raw/weeks/ and {vault}/raw/artifacts/.
  • Boundary markers: Absent; the output is structured via references/recall-output-template.md but lacks explicit boundary instructions to the model.
  • Capability inventory: The script performs file reads and JSON parsing; it does not have network or dangerous execution capabilities.
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
May 12, 2026, 09:56 AM
Security Audit — agent-trust-hub — lode-session-start-recall