The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several Python scripts (e.g., discover_skills.py, package_skill.py, triage_skill_request.py, validate-skill.py) intended for local execution. These scripts use standard Python libraries to perform file system operations such as indexing available skills, reading markdown files, and creating zip archives. These operations are scoped to the user's local environment (~/.claude/ and ~/.cache/) and are consistent with the skill's stated purpose as a developer tool.- [DATA_EXFILTRATION]: While the skill reads local directory structures to index and package skills, it lacks any network-capable code (e.g., requests, socket, urllib) to transmit this data to external servers. All operations are local.- [PROMPT_INJECTION]: The skill uses natural language triggers and iterative analysis phases to guide the AI's behavior. No adversarial instructions designed to bypass the agent's core safety filters or extract system prompts were found.- [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code execution. It generates local scripts from predefined templates in a deterministic manner for the user to verify.- [SAFE]: The skill utilizes well-structured Python code following standard patterns, such as argparse for CLI interfaces and dataclasses for result handling. It includes a multi-agent synthesis protocol that acts as an internal quality control mechanism for any generated content.

skillforge