The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of test code within the workspace using the mcp__lsp__run_tests tool and analyzes results using the grep command. While this is the intended purpose for test correlation, it involves running code defined in the project files.
[PROMPT_INJECTION]: The skill processes data from the workspace and test outputs, creating a surface for indirect prompt injection.
Ingestion points: Workspace symbol names from mcp__lsp__find_symbol and test execution output from mcp__lsp__run_tests.
Boundary markers: No delimiters or instructions are used to separate untrusted tool output from the agent's logic.
Capability inventory: The agent can execute tests and commands, and has read access to the workspace.
Sanitization: There is no evidence of sanitization or filtering of external content before it is processed and summarized for the user.
[EXTERNAL_DOWNLOADS]: The skill references the agent-lsp repository on GitHub as a required dependency, which is maintained by the skill's author.

lsp-test-correlation