skills/lilmgenius/paperthin/backprop/Gen Agent Trust Hub

backprop

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes 'output' from previous agent cycles (such as corrections or failed cycles) to update source files, creating an indirect prompt injection surface.
  • Ingestion points: The skill ingests 'output the lesson came from' in Step 1 to identify necessary changes.
  • Boundary markers: The instructions do not specify delimiters or instructions to treat output as untrusted.
  • Capability inventory: The skill utilizes several tools (re0-memo, elon, ssotize, re0, debloat, re0-order) capable of modifying repository documentation and source code.
  • Sanitization: No sanitization of input is described, though user approval gates in Step 3 and Step 4 serve as a mitigation.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 04:52 AM