The Agent Skills Directory

[SAFE]: The skill defines a six-phase methodology for systematic debugging, focusing on creating reliable feedback loops and structured hypothesis testing. Analysis of the instructions and references found no malicious patterns.
[EXTERNAL_DOWNLOADS]: The skill references standard project files (.agent-skills/skill-standardization/SKILL.md) and a local validator script (.agent-skills/skill-standardization/scripts/validate_skill.sh). These are internal project references and do not involve untrusted remote sources.
[COMMAND_EXECUTION]: The skill provides examples of standard development commands like npm test and console.error for diagnostic purposes. These are expected in a debugging skill and do not involve privilege escalation or obfuscated execution.
[PROMPT_INJECTION]: No prompt injection attempts, such as 'ignore previous instructions' or safety filter bypasses, were detected in the skill text or metadata.
[DATA_EXFILTRATION]: No patterns indicative of sensitive data access or external transmission were found. The tool access list (Read, Grep, Glob, Bash, Write, Edit) is appropriate for the stated purpose of code diagnosis.

diagnose