qstack-be-concise
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a utility for rewriting text to be more concise and non-technical. It contains no executable scripts, shell commands, or network activity.
- [INDIRECT_PROMPT_INJECTION]: The skill processes the assistant's conversation history, which is a potential ingestion point for instructions embedded in external data previously read by the agent. The risk is minimized because the skill lacks the ability to execute commands or interact with external systems.
- Ingestion points: The skill processes the immediately previous assistant response.
- Boundary markers: No specific delimiters or boundary markers are defined for the input text.
- Capability inventory: The skill has zero functional capabilities; it does not invoke tools, access the network, or interact with the file system.
- Sanitization: The instructions mandate the removal of jargon, implementation details, file paths, and command names, which serves as a content-filtering mechanism.
Audit Metadata