The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes Python scripts that interact with the local filesystem to analyze and create files.
scripts/complexity_check.py reads file content provided via command-line arguments to calculate complexity metrics.
scripts/scaffold_test.py creates new directories and files on the filesystem based on provided source file paths without strict path validation or sanitization.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process, analyze, and refactor external source code, creating a surface for embedded instructions to influence the agent.
Ingestion points: scripts/complexity_check.py reads file content; the agent is instructed to read src/ files during the Phase 1 (Design & Plan) and Phase 3 (Review & Refine) workflows.
Boundary markers: No explicit delimiters or 'ignore' instructions are used when reading external code to differentiate code from potential instructions.
Capability inventory: The skill includes scripts for file writing (scaffold_test.py) and instructions for the agent to modify application source code.
Sanitization: No sanitization is performed on the content of processed files before they are analyzed or used to generate new file structures.
[DATA_EXFILTRATION]: The skill files contain hardcoded absolute paths that reveal information about the author's local directory structure.
SKILL.md contains multiple references to /Users/sargupta/SahayakAIV2/..., exposing the username and internal project layout of the developer.

senior-software-engineer