The Agent Skills Directory

[SAFE] (SAFE): No executable code detected. The skill consists exclusively of Markdown files (.md) containing text-based instructions, tables, and documentation templates.- [SAFE] (SAFE): Standard methodologies. The investigation techniques described (Binary Search, Variable Isolation, Process of Elimination) are standard industry practices for incident response and debugging.- [SAFE] (SAFE): Safe examples. Code blocks in the documentation, such as the git bisect example in references/elimination-techniques.md, are clearly marked as illustrative examples and do not instruct the agent to download or execute untrusted scripts from the internet.- [SAFE] (SAFE): Indirect Prompt Injection Surface. While the skill is designed to process external data (logs and incident reports) which is a known vector for indirect prompt injection, the skill itself provides structured templates and logical frameworks that encourage systematic analysis rather than blind execution of data-driven commands. Following the security framework, this surface is noted but considered SAFE given the purely instructional nature of the skill contents.

find-root-causes