fulcrum
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to check the FULCRUM_FEED_LOG_DIR environment variable and perform git commits for each file modification.
- [INDIRECT_PROMPT_INJECTION]: The skill processes text from voice memo exports, creating a surface for potential injection from untrusted input.
- Ingestion points: Voice memo export documents containing feeding records.
- Boundary markers: Uses a specific sentinel phrase 'For Fulcrum' to trigger processing logic.
- Capability inventory: Performs file system reads, writes, and moves within a local Obsidian vault, along with git commit operations.
- Sanitization: Implements normalization for weight units and ingredient blend names.
Audit Metadata