reality-audit
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a structured framework for the agent to execute shell commands for building, packaging, deploying, and testing software deliverables. This protocol is intended to verify implementation accuracy through real-world terminal output and exit codes. Evidence: Section 3 ('Прогони терминальную лестницу свидетельств') describes steps to build and test canonical deliverables.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon external specifications, requirements, and API contracts to generate its verification assertions. This introduces a surface where malicious instructions within those documents could attempt to manipulate the auditing logic.
- Ingestion points: Ingests requirements, specifications, and owner-provided corrections (SKILL.md, Section 1).
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded prompts within the source documents.
- Capability inventory: The agent is directed to perform build operations, test execution, and deployment tasks in the terminal (SKILL.md, Section 3).
- Sanitization: There are no explicit requirements for sanitizing or escaping content from the ingested specifications before assertions are formed.
Audit Metadata