encode-invariant

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from repository files to generate validation logic. Ingestion points: AGENTS.md, docs/WORKFLOW.md, and local repository files. Boundary markers: Absent. Capability inventory: Writing validation files and executing repository-native tools (lint, test, build). Sanitization: Not explicitly defined in instructions.
  • [COMMAND_EXECUTION]: The skill instructs the agent to run repository-native tasks, including tests, builds, and linters, to provide proof of validation.
  • [DYNAMIC_EXECUTION]: The skill involves generating and running scripts or test mutations to verify repository invariants.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 08:59 AM
Security Audit — agent-trust-hub — encode-invariant