docs: Add AGENTS.md conventions for coding agents - #235
Conversation
Yep, this sets up the "persona" and basic backgrounder files. We can also add a first "How can I help you get started?" type response back to the user with a path of things it should know how to help with:
|
Give AI coding agents standing instructions for working in this repository: grounding requirements, analysis and verification expectations, answer format and voice, and scope limits. CLAUDE.md points Claude Code at AGENTS.md so both tools share one source. Signed-off-by: Michael Scott <mike@foundries.io> Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Collect the writing conventions where human contributors can find them: the Markdown doc rules established in the PR#238 review round, plus the code comment, commit message, and trailer conventions that AGENTS.md directs coding agents to follow. Signed-off-by: Michael Scott <mike@foundries.io> Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
7c9b4aa to
b394461
Compare
|
Added a STYLE.md based on findings in #238 |
|
Oof. I realize, I should have used: https://github.com/foundriesio/fio-style/blob/main/style-guide.md |
No worries. In the near future, the style guide will be downloaded with the rest of fio-style—along with a set of skills and linter rules—for all our repos. |
|
@kprosise Should we wait for the work you're doing on fiostyle? |
Up to you. On the one hand I don't have Agent specific skills set, will need to setup everything to make it easier to use and semi-automated, and this will take a little bit…on the other hand, you could potentially point the agent to the style guide/merge this PR now while I work on finishing fio-style stuff, and then we can change it up later. |
Give AI coding agents standing instructions for working in this repository: grounding requirements, analysis and verification expectations, answer format and voice, scope limits, and commit conventions. CLAUDE.md points Claude Code at AGENTS.md so both tools share one source.