feature-development
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interpolates raw user input from the
$ARGUMENTSplaceholder directly into the prompts for several specialized sub-agents, such as the architect and implementation SMEs. - Ingestion points: The
$ARGUMENTSvariable inSKILL.mdacts as a direct entry point for untrusted user input into the agent's workflow. - Boundary markers: The skill does not use delimiters (e.g., XML tags, triple backticks) or specific instructions to the sub-agents to ignore potentially malicious content within the feature description.
- Capability inventory: The sub-agents are tasked with modifying code, generating tests, and interacting with the CLI to manage Hosted Clusters. These powerful capabilities could be misdirected if the input contains adversarial instructions.
- Sanitization: There is no evidence of input validation, escaping, or filtering to prevent the user input from overriding the sub-agents' primary instructions.
Audit Metadata