The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill utilizes bash, grep, and dotnet CLI tools for legitimate development purposes such as codebase exploration and database migration generation. All commands are targeted at local project directories and follow standard patterns.
[Indirect Prompt Injection] (LOW): This skill exhibits a surface for indirect prompt injection. 1. Ingestion points: Untrusted data enters the context via grep and Read tools during Phase 1 (Discovery). 2. Boundary markers: Explicit delimiters or 'ignore' warnings for code comments are absent in the provided prompts. 3. Capability inventory: The skill possesses Write, Edit, and Bash capabilities which could be abused if the agent follows instructions hidden in code comments. 4. Sanitization: No explicit sanitization or filtering of codebase content is defined. Mitigation: The risk is significantly reduced by the mandatory human-in-the-loop 'Approval Gate' (Phase 4) and explicit 'Evidence Chain Validation' protocols.

tasks-feature-implementation