elon
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of requirements, plans, or specifications, creating a surface for indirect prompt injection attacks where malicious instructions could be embedded in the data. \n
- Ingestion points: Requirement lists or project commitments provided by the user in SKILL.md (Workflow Step 1). \n
- Boundary markers: Data is structured within a Markdown table with specific column headers (Workflow Step 9). \n
- Capability inventory: The skill executes local Node.js scripts to validate tables and ledgers, and the agent is instructed to append entries to a local JSONL file. \n
- Sanitization: The provided scripts validate the structure (JSON/Markdown table) but do not perform sanitization or filtering of the text content for embedded prompt instructions. \n- [COMMAND_EXECUTION]: The skill invokes local utility scripts (scripts/table.mjs and scripts/ledger.mjs) using Node.js to validate input formatting and generate audit reports.
Audit Metadata