The Agent Skills Directory

[SAFE]: The skill consists entirely of natural language instructions designed to guide a human user through an observation process. It does not contain any executable scripts, binary files, or automated commands.
[SAFE]: No prompt injection or behavior override patterns were detected. The instructions are pedagogical and align with the stated purpose of coaching neutral attention and debugging.
[SAFE]: There are no network operations, hardcoded credentials, or sensitive file path accesses. While it mentions observing codebases or logs, it directs the user to perform these actions and record notes manually rather than automating data collection.
[SAFE]: No obfuscation (Base64, zero-width characters, etc.) or hidden content was found within the text or metadata.
[SAFE]: The skill only allows the 'Read' tool, which is used for its intended purpose of processing the instruction file itself or user-provided data, posing no privilege escalation or RCE risks.

observe-guidance