[AIAA] Add repository instructions for AI agents - #369
Conversation
Assisted-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: Nate W <natew@cncf.io>
✅ Deploy Preview for cncf-techdocs ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
There was a problem hiding this comment.
Pull request overview
Adds repository-wide guidance for AI agents, covering safety, evidence standards, methodology references, and validation.
Changes:
- Defines AI-agent ground rules and evidence requirements.
- Directs agents to existing assessment methodology.
- Documents repository checks.
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
Assisted-by: Copilot <223556219+Copilot@users.noreply.github.com> Signed-off-by: Nate W <natew@cncf.io>
|
@nate-double-u - Should the copilot-instructions include writing style instructions or should they be handled at a more granular level like in prompt-related files? Such style instructions include regarding verbosity, handling parenthetical clauses, and a few others. |
|
I think granular. This file loads into every Copilot surface in the repo (including code review on config PRs), so it should only list what applies to all agent work: safety, evidence, pointers. I think that style for assessment deliverables should be a part of the the per-phase agent profiles that'll get built in later steps. If a style rule turns out to be needed repo-wide, we can add it here then. |
github-project-automation
Bot
moved this to Todo
in AI-assisted TechDocs Assessment Build
nate-double-u
moved this from Todo
to In Progress
in AI-assisted TechDocs Assessment Build
|
|
||
| ## Point at the methodology, do not copy it | ||
|
|
||
| Read criteria, process, and templates from `docs/analysis/` and refer to them by |
There was a problem hiding this comment.
Perhaps add: "Do not offer analysis devoid of the criteria described."
|
|
||
| ## Checks | ||
|
|
||
| Run `npm run check` before proposing changes: it covers formatting, markdown |
There was a problem hiding this comment.
Suggest adding: "Avoid introducing to CSpell terms, initialisms, or acronyms that can be otherwise be understood by being written out."
Adds
.github/copilot-instructions.md: ground rules for any AI agentworking in this repository. Build step 2 of the assessment system
(spec section 16); reviewed against sections 8 and 14. Tracking: #368.
What it carries
stay in this repository; a human reviews and merges every change;
secrets stay out of outputs.
reproducible numbers, cited findings, unverifiable content flagged,
not asserted.
docs/analysis/, never restate it (P-2).npm run check.Shape choices
including code review on ordinary PRs. Nothing phase-specific; that
belongs to the agent profiles (later step).
is self-sufficient; the spec stays a design artifact.
Verification
(required rules present, no spec references, length bounds, style).
npm run checkpasses.