agentleFS
aniq-ui-proaniq-ui-pro/agent-skills/skills/add-admin-dashboard/SKILL.md

add-admin-dashboard

Skill1 starsChanged 2 days agoMITView on GitHub

Connect an existing Next.js admin dashboard to an existing API. Use when asked to add an admin panel, back office or staff dashboard to an app that already has an API (NestJS controllers, Next.js route handlers, or an OpenAPI file). Maps the API's resources, adds a client module and query hooks per resource, wires sign-in and roles, adds one list page and one form per resource, and verifies with a type check, a build and real requests.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run node ${CLAUDE_SKILL_DIR}/scripts/map-api.mjs <api-folder-or-openapi-file> and save the output to docs/admin-api-map.txt.

    line 14
  2. Distinctive:

    Check: each chosen resource has a list route and a write route in the map, and every route under "routes with no guard" is meant to be …

    line 18
  3. Distinctive:

    Check: login with curl returns a token.

    line 37
  4. Very specific:

    Check: npx tsc --noEmit passes with no new errors.

    line 46
  5. Distinctive:

    Use the dashboard's table component.

    line 50
  6. Distinctive:

    Check: signed in, the page shows the same total as the list route returns to curl.

    line 54
  7. Distinctive:

    Check: an empty required field shows an error and sends no request.

    line 62
  8. Very specific:

    Run the type check, lint and next build.

    line 67
  9. Distinctive:

    Check: the build passes and both bad writes answer 400.

    line 70
  10. Very specific:

    No secret in a NEXT_PUBLIC_ variable: the browser can read every one of them.

    line 74
  11. Distinctive:

    Add one resource, check it, then start the next.

    line 76

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Add an admin dashboard to an existing API

Work in the dashboard's own conventions. Before writing anything, find one resource it already handles end to end (API function, query hook, list page, form) and copy that shape. Components never call the API directly, and there is only one HTTP client.

1. Map the API

Run node ${CLAUDE_SKILL_DIR}/scripts/map-api.mjs <api-folder-or-openapi-file> and save the output to docs/admin-api-map.txt. Agree with the user which resources to manage: usually the ones marked list page + ... form.

Check: each chosen resource has a list route and a write route in the map, and every route under "routes with no guard" is meant to be public, or reported to the user before going on.

2. Point the dashboard at the API

Set the API base URL in the dashboard's env file (the variable its HTTP client already reads; NEXT_PUBLIC_ values are baked in at build time, so restart or rebuild after). Add the dashboard's origin to the API's CORS allow-list.

Check: curl -s -o /dev/null -D - -X OPTIONS "<api>/<list route>" -H "Origin: <dashboard origin>" -H "Access-Control-Request-Method: GET" prints Access-Control-Allow-Origin: <dashboard origin>.

3. Wire sign-in and roles

Make the dashboard's sign-in call the API's login route and map its answer (token, user, roles, permissions) into what the dashboard already stores. The HTTP client sends the token; a 401 clears the session and opens the sign-in page. Gate buttons with the permission names from the map.

Check: login with curl returns a token; one list route answers 401 without it and 200 with it; a user without the write permission gets 403 from that write route.

4. Client layer and query hooks, one resource at a time

One API module per resource (list, get, create, update, delete), typed from the request bodies in the map. Then query keys, a list query that passes the route's pagination and filter params, and mutations that invalidate the resource's keys.

Check: npx tsc --noEmit passes with no new errors.

5. One list page per resource

Use the dashboard's table component. Columns come from the response fields, pagination and filters from the list route's query params. Show create, edit and delete only to users with those permissions. Add the page to the navigation.

Check: signed in, the page shows the same total as the list route returns to curl.

6. One form per resource

Fields are the create body's fields; required ones are marked * in the map. Validate with the form library the dashboard already uses (react-hook-form with zod if it has none) and show the API's 400 field errors on the matching fields.

Check: an empty required field shows an error and sends no request; a valid submit answers 201 and the new row appears in the list.

7. Verify

Run the type check, lint and next build. Repeat step 3's curl checks for every resource, plus one write with an unknown field and one with a required field missing.

Check: the build passes and both bad writes answer 400. A 500 is an API bug: report it.

Rules

  • No secret in a NEXT_PUBLIC_ variable: the browser can read every one of them.
  • Hiding a button is not access control. The API must answer 403 as well.
  • Add one resource, check it, then start the next.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.