agentleFS
redhat-developerredhat-developer/rhdh-skills/skills/plugins/rhdh-plugin-authoring/SKILL.md

rhdh-plugin-authoring

Skill20 starsChanged 20 days agoApache-2.0View on GitHub

Writes Backstage plugin source for Red Hat Developer Hub in rhdh-plugins or community-plugins: scaffold a backend dynamic plugin with createBackendPlugin or createBackendModule from @backstage/backend-plugin-api, scaffold a frontend dynamic plugin, and implement features against Backstage UI (@backstage/ui), MUI, Scalprum bundles, fetchApi, i18n, and the Backstage test utilities. Use for "scaffold a backend dynamic plugin", "create a frontend dynamic plugin", "implement this plugin feature", entity cards, scaffolder actions, catalog processors, backend extension points, renderInTestApp and TestApiProvider tests, dev app setup, and the local build gates yarn tsc:full and yarn build:api-reports:only.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the target repository's AGENTS.md or CLAUDE.md and any linked specification or issue.

    line 26
  2. Very specific:

    Run the workspace build gates in order and stop at the first failure: yarn prettier:fix, yarn tsc:full, yarn build:all, yarn test …

    line 95
  3. Distinctive:

    Do not stage, commit, push, or open a pull request here.

    line 98

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

RHDH Plugin Authoring

Own plugin source code: creating a new Backstage dynamic plugin for RHDH and implementing features inside an existing one. Backend and frontend are branches of the same job — the same conventions, the same build gates, the same repo — so choose the branch from the request rather than asking for a skill.

Start here

  1. Read the target repository's AGENTS.md or CLAUDE.md and any linked specification or issue. Repository rules beat anything written here.
  2. Run python scripts/detect-rhdh-context.py --path <plugin-or-workspace> in an existing checkout. Record role, frontend system, plugin ID, dynamic status, MUI version, and package manager.
  3. Establish the target RHDH and Backstage versions. Prefer an explicit user or repository value; otherwise invoke /rhdh-context for repository, tool, and version facts. If it is unavailable, ask the user instead of guessing.
  4. Inspect branch and status before modifying files, and protect uncommitted work.

For a new plugin or a change to a public API where materially different designs are valid and the request does not settle the choice, invoke /grilling as a design gate, use the constraints it produces, and show the design before implementing. If /grilling is not installed, say so, name /setup-rhdh-skills install as the human's next step, and pause that branch.

Route by outcome

OutcomeLoad and follow
Scaffold a backend dynamic pluginreferences/backend.md; run scripts/scaffold.py
Scaffold a frontend dynamic pluginreferences/frontend.md; run scripts/scaffold.py
Decide what kind of extension to buildreferences/plugin-types.md
Implement or review plugin codereferences/development-patterns.md, then only the relevant reference below
Style with Backstage UI or MUIreferences/bui.md
Write NFS code in a new pluginreferences/nfs.md
Apply RHDH-specific patternsreferences/rhdh.md
Run or extend the plugin dev appreferences/dev-app.md
Write tests for plugin codereferences/testing.md
Write or read a frontend feature specreferences/frontend-specs.md

Infer the branch from the request. Ask only when the missing choice changes the implementation, such as backend versus frontend.

Boundaries

  • This skill writes plugin source. It does not export artifacts, generate deployment configuration, or publish anything.
  • /rhdh-plugin-export owns exporting and packaging a plugin for RHDH.
  • /rhdh-plugin-wiring owns generating dynamic-plugins.yaml configuration and mount points for a finished frontend plugin. Reach for it by name rather than writing wiring configuration here.
  • /rhdh-plugin-nfs-migration owns converting an existing legacy frontend plugin to the New Frontend System. This skill writes NFS code in a plugin that is already on it.
  • /rhdh-backstage-upgrade owns moving @backstage/* dependency versions.
  • /rhdh-test-placement advises which repository and layer a permanent test belongs in, before you write it here.
  • /rhdh-pr-create owns changesets, staging, signed-off commits, and pull requests. Leave changed files unstaged and list them when you finish.

Invoke a named skill and describe the handoff in the conversation. Never open another skill's files.

Invariants

  • Backend code uses the new backend system only — createBackendPlugin or createBackendModule from @backstage/backend-plugin-api. Never the legacy backend system.
  • A default export from src/index.ts is required. Its absence is the most common cause of a dynamic plugin that will not load in RHDH.
  • The Scalprum name must match the key used in dynamic-plugins.yaml wiring.
  • Match the neighbouring implementation: check two or three sibling workspaces before inventing a configuration, fixture path, or utility pattern.
  • Build incrementally. Write one component or hook, type check, and verify a visual change in a running app before writing the next.
  • Mirror a current neighbouring test rather than recalling a Backstage test API from memory; that surface moves between releases.
  • Run the workspace build gates in order and stop at the first failure: yarn prettier:fix, yarn tsc:full, yarn build:all, yarn test --watchAll=false, yarn build:api-reports:only.
  • Do not stage, commit, push, or open a pull request here.

Completion

Report the branch taken, the files created or changed and left unstaged, the build gates run and their results, any public API surface added, the remaining risks, and the named skill to invoke next for wiring, export, testing, or publication.

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.

Reviews can't be read right now.

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.