The Agent Skills Directory

Command Execution: The skill's workflow instructs the agent to identify and execute test commands defined in project manifests (e.g., dotnet run) to establish a baseline. This involves running arbitrary code provided by the project being analyzed, which is a consideration when the repository's contents are not fully trusted.
Indirect Prompt Injection Surface: The skill relies on processing external production code, test files, and manifests to perform its gap analysis. This ingestion of untrusted data creates a surface where malicious instructions hidden in comments or code could attempt to manipulate the agent's logic. (1) Ingestion points: Production source files, test suites, and project manifests accessed in SKILL.md. (2) Boundary markers: The instructions do not explicitly mandate the use of delimiters or 'ignore' instructions when reading the external code into the agent context. (3) Capability inventory: The skill utilizes file system discovery, file reading, and shell command execution. (4) Sanitization: There is no specific process mentioned for sanitizing or escaping the content of the files before they are interpreted by the agent.

test-gap-analysis