replica / coreai
AGENTS.md117 starsChanged 10 days agoMITView on GitHub
Read the shared Replica instructions before prototype work. They own task scope, component preflight, canvas setup, publishing and browser validation. This file and the CoreAI skills own app-specific conventions. The CoreAI Prototype is a React Router single-page app for exploring agent and design-system experiences with local mock data. Run commands from this app's directory: The app runs at <http://127.0.0.1:5174>. Vitest runs hook tests in jsdom with React Testing Library. Keep the chat hook real and use simulated timers to test…
We split this AGENTS.md into the rules it gives an agent
12 rules. About 1 is common, 2 are distinctive, and 9 are very specific to this project.
≈1
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
9
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Very specific: line 3
Read the shared Replica instructions before prototype work.
- Distinctive: line 19
Run commands from this app's directory:
- Very specific: line 58
When more than one row applies, load each relevant skill.
- Very specific: line 66
Import product-facing components from @coreai-ds/react-components.
- Very specific: line 67
Import app-owned adapters from ~/components/common.
- Very specific: line 68
Import Fluent icons from @fluentui/react-icons.
- Very specific: line 69
Do not import components directly from @fluentui/react-components.
- Very specific: line 70
Keep component styles in colocated *.module.css files.
- Very specific: line 72
Put mock records in app/data/.
- Distinctive: line 75
Keep copyright and MIT license headers on source files.
- Very specific: line 76
Do not commit generated build/, dist/, or .react-router/ output.
- Common: line 80
Keep chat mock-only. Do not add API-key settings or live-model requests.
≈ “Common” is an estimate: at least 1 in 5 of the 36,368 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CoreAI Prototype — AI Coding Assistant Instructions
Read the shared Replica instructions before prototype work. They own task scope, component preflight, canvas setup, publishing and browser validation. This file and the CoreAI skills own app-specific conventions.
The CoreAI Prototype is a React Router single-page app for exploring agent and design-system experiences with local mock data.
Project context
- React 18, React Router v7 framework mode, Vite, and TypeScript
- Client-side rendering only (
ssr: false) ~/resolves toapp/- CSS Modules for component and page styles
- TanStack Query over mock JSON in
app/data/ - URL-driven prototype features and design variants
- Simulated chat over local canned replies, without credentials or model requests
Run and validate
Run commands from this app's directory:
npm ci
npm run devThe app runs at http://127.0.0.1:5174.
npm run type-check
npm run lint
npm test
npm run build
npm run previewVitest runs hook tests in jsdom with React Testing Library. Keep the chat hook real and use simulated timers to test reply cycling and cancellation. Tests must not contact model services or use real credentials. Run a focused file with npm test -- app/hooks/useChat.test.ts. Verify visual changes in a browser.
Skill index
Detailed working conventions live in .agents/skills/. Load the matching skill before
making a change. Treat the skill as the source of truth instead of duplicating its
instructions here.
| Task or area | Skill |
| Start or open the app | start-app |
| Format, lint, or type-check | lint |
| Choose CoreAI components, APIs, templates, or patterns | coreai-ds-fluentui-react |
| Edit TSX or component imports | react-components |
| Edit CSS Modules or debug style precedence | css-styling |
| Add or change icons | icons |
| Add routes or change navigation | routing |
| Change query hooks, mock data, or data state | state-management |
| Configure or consume features and variants | features-variants |
| Create a new design alternative | new-variant |
| Interpret rendered HTML or generated class names | interpreting-html |
When more than one row applies, load each relevant skill. In particular, component work
often requires both coreai-ds-fluentui-react and react-components, plus a styling,
icon, routing, state, or variant skill when applicable.
For design-system reference, follow the coreai-ds-fluentui-react skill from the CoreAI documentation index.
Baseline boundaries
- Import product-facing components from
@coreai-ds/react-components. - Import app-owned adapters from
~/components/common. - Import Fluent icons from
@fluentui/react-icons. - Do not import components directly from
@fluentui/react-components. - Keep component styles in colocated
*.module.cssfiles. - Register routes in
app/routes.ts. - Put mock records in
app/data/. - Register prototype options in
app/features.json; do not build a second feature system. - Keep copyright and MIT license headers on source files.
- Do not commit generated
build/,dist/, or.react-router/output.
Data and credentials
Keep chat mock-only. Do not add API-key settings or live-model requests. Do not commit credentials, endpoints, real customer data, prompts, or responses.
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 28,693 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- hub/AGENTS.mdAGENTS.md
- starters/minimal/AGENTS.mdAGENTS.md
- replica-pick-and-polishSkill
- replica-publish-prototypeSkill
- review-hub-feedbackSkill
- code-reviewSkill
- write-testsSkill
- coreai-ds-fluentui-reactSkill
- css-stylingSkill
- features-variantsSkill
- iconsSkill
- interpreting-htmlSkill
- lintSkill
- new-variantSkill
- react-componentsSkill
- routingSkill
- start-appSkill
- state-managementSkill
- context-templatesSkill
- accessibility-auditSkill
- add-seed-dataSkill
- agent-browserSkill
- browser-canvas-validationSkill
- component-design-system-reviewSkill
- create-prototypeSkill
- developer-handoffSkill
- prepare-feedback-proposalsSkill
- product-designSkill
- prototype-controlsSkill
- responsive-auditSkill
- review-hub-feedbackSkill
- setup-skillsSkill
- visual-design-reviewSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

