Write Adr

What to consider before installing: - The skill will run git and python commands and will read and write files in your repository (it looks for docs/adrs/ and writes to docs/adr/ — note the plural vs singular inconsistency). Confirm you are comfortable with an agent that can inspect repo files and create ADR markdown files. - The SKILL.md instructs executing a local script (skills/adr-writing/scripts/next_adr_number.py). If that script exists in the repo it will be executed — review its contents before running to ensure it's safe. - The skill loads and invokes other named skills/subagents (beagle-analysis:adr-decision-extraction and beagle-analysis:adr-writing). Verify those skills exist and are trusted; they may have their own requirements or behaviors. - The metadata does not declare required binaries (git, python). Ensure those tools are available and consider testing in a safe environment or on a copy of the repo first. - The disable-model-invocation flag is set but the instructions expect subagent model calls; confirm how the platform enforces this and whether the skill will function as intended. - If you need higher confidence: ask the publisher for the source (or a trusted repository), a clear list of required binaries, and corrected paths (docs/adrs vs docs/adr). If the author provides those and the referenced sub-skills are trusted, the skill would be coherent and lower-risk.