The Agent Skills Directory

[SAFE]: The skill consists of documentation and code examples intended to guide an AI agent in using specialized tools for context optimization. It does not contain executable code itself, only patterns for the agent to follow.
[COMMAND_EXECUTION]: The skill explicitly documents the use of execution tools (ctx_execute, ctx_execute_file) for running JavaScript, Python, and Shell code. These are intended for data analysis, log parsing, and infrastructure inspection within a sandboxed environment.
[EXTERNAL_DOWNLOADS]: Examples within the skill demonstrate how to fetch data from external APIs (e.g., GitHub API) and web documentation. These examples use well-known domains and are consistent with the skill's stated purpose of processing API responses and indexing documentation.
[DATA_EXFILTRATION]: There is no evidence of unauthorized data transfer. The patterns provided focus on local processing and summarization, with network activity limited to standard development workflows (API calls to known services).
[PROMPT_INJECTION]: The skill uses strongly worded instructions (e.g., 'MANDATORY RULE', 'CRITICAL RULE') to enforce token-saving behaviors. These do not target the agent's safety guidelines or attempt to override system prompts for malicious purposes.

context-mode