adaptixc2-dev
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill guides the agent to perform standard development tasks using established tools such as
go,git,make, andrgfor auditing, building, and testing components. These operations are transparent and scoped to the user's development environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill assists in the review of third-party source code (extenders and scripts), which introduces a theoretical surface for indirect prompt injection if the analyzed code contains malicious instructions.\n - Ingestion points: The agent reads and processes extender source code and AxScript UI definitions as part of development and review tasks.\n
- Boundary markers: While the documentation stresses rigorous manual verification and validation of code boundaries, it does not define specific prompt delimiters for the agent's internal ingestion process.\n
- Capability inventory: The agent leverages local shell tools (
go,make,axtool) to interact with the project codebase as specified in the development workflow.\n - Sanitization: The instructions provide extensive guidance on sanitizing user-provided data within the built extensions but do not specify automated sanitization for the code the agent is reviewing.
Audit Metadata