The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from the local filesystem in the form of C# source code and project files, creating a surface for instructions embedded in code to influence agent behavior.
Ingestion points: The skill reads source code via grep for TODOs (Step 5) and uses various MCP tools such as get_diagnostics and find_dead_code to analyze the project structure and content.
Boundary markers: The instructions do not define delimiters (e.g., XML tags or markdown blocks) to isolate content read from the codebase from the skill's own operational instructions.
Capability inventory: The skill possesses the capability to execute shell commands including dotnet build and dotnet test, which involve compiling and running code defined within the processed project files.
Sanitization: No explicit sanitization or validation of the ingested code content is performed before it is presented to the agent or passed to execution tools.

de-sloppify