The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/orchestrate.py uses the subprocess.run function to execute sibling scripts within the communication skill suite (analyze_file.py and summarize_folder.py). The implementation uses sys.executable and an argument list, which is a secure pattern that avoids shell injection vulnerabilities. The executed scripts are local and part of the same vendor suite.- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided directory paths and the content of all markdown files within those directories, creating a surface for indirect prompt injection.
Ingestion points: The target_path argument in scripts/orchestrate.py and the content of recursively discovered .md files.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the orchestration logic.
Capability inventory: The script has the capability to execute other Python processes via subprocess.run based on the discovery of files.
Sanitization: While the script validates path existence and file extensions, it does not sanitize or validate the internal content of the markdown files before passing them to analysis tools.

communication-orchestrator