The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute standard development CLI tools and project-local scripts to audit code and verify the integrity of refactors.\n
Evidence: Uses git log, Get-ChildItem, and dotnet CLI for repository analysis in SKILL.md.\n
Evidence: Directs the agent to run project-specific scripts such as ./build.ps1, tools/selftest.ps1, and tools/snap.ps1 to prove behavior remains unchanged during refactoring.\n- [INDIRECT_PROMPT_INJECTION]: The skill analyzes documentation and source files within the workspace, which represents an attack surface where maliciously crafted content could influence the agent's refactoring proposals.\n
Ingestion points: Reads workspace files including CLAUDE.md, ARCHITECTURE.md, docs/adr/, docs/FIXLOG.md, and C# source files.\n
Boundary markers: No explicit delimiters or instructions to ignore embedded prompts are provided during the survey and analysis phase.\n
Capability inventory: The agent has permissions to modify local files, execute shell commands, and run build/test scripts.\n
Sanitization: No sanitization or schema validation is applied to the content read from workspace files before processing.

improve-architecture