The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/symbol_map.py executes git and find commands to locate source files within the project directory.
Evidence: Calls to subprocess.run occur in the run_command function within scripts/symbol_map.py.
Context: The commands are executed as lists without a shell environment (shell=False), which prevents shell injection, and patterns are derived from internal configuration.
[PROMPT_INJECTION]: The skill exhibits a vulnerability surface for indirect prompt injection (Category 8) due to its processing of untrusted project data.
Ingestion points: The script scripts/symbol_map.py reads the contents of various source files (TS, JS, Python, Rust, Go) to extract symbol names.
Boundary markers: While the extracted data is presented in Markdown tables, there are no instructions for the agent to ignore or sanitize instructions that might be embedded within symbol names (e.g., a function name containing injection patterns).
Capability inventory: The skill performs file system reads, writes to a local maps directory, and executes discovery commands (git, find).
Sanitization: No sanitization or escaping is performed on the extracted symbol names before they are interpolated into the final markdown report.

map