agentleFS
ack-solutionsack-solutions/claude-code-ai-skills/skills/feature-delivery/SKILL.md

feature-delivery

Skill1 starsChanged 44 days agoMITView on GitHub

Implement a defined feature across affected UI, API, data and tests. Use for build this feature, implement the approved flow, or wire this UI to the API. Planning-only requests belong to product-planning; audit-only requests belong to the relevant review workflow.

8 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read applicable project instructions and the relevant acceptance criteria.

    line 12
  2. Distinctive:

    Put business rules in their existing domain owner.

    line 20
  3. Distinctive:

    Keep technical and product documentation aligned when behaviour changes.

    line 25

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “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

Feature delivery

Deliver the requested user outcome through the relevant layers of the existing project. Explicit user instructions take precedence over this workflow's defaults.

Scope and evidence

Read applicable project instructions and the relevant acceptance criteria. Inspect the manifests, nearby implementation, contracts and tests to learn the actual stack and supported commands. Preserve unrelated changes. In a new project, establish only the foundations needed for the requested feature.

Trace affected consumers before changing shared behaviour. For substantial work, make a short implementation plan covering dependencies, data impact and verification. Resolve material requirement conflicts; use disclosed reasonable defaults for reversible implementation details. Do not invoke every skill or demand a new specification for a small edit.

If the full fix requires edits outside the authorized task, explain the impact and obtain expanded scope before those edits. Keep unresolved decisions and their owners in the project's existing records; only material dependencies block the affected work. Use its adopted task lifecycle rather than adding parallel statuses.

Implement

  • Put business rules in their existing domain owner. Keep transport, persistence and presentation responsibilities clear without inventing layers solely to fit a pattern.
  • Reuse authoritative contracts, utilities, design tokens and components. Extend them when the requirement warrants it; avoid parallel copies that drift.
  • Handle relevant validation, ownership, permissions, partial failures, retries, concurrency and recovery. A hidden UI control is not authorization.
  • Preserve compatibility with existing stored data and supported clients. Schema changes follow the project's migration process; inspect generated migrations.
  • Connect UI actions to real outcomes within scope. Loading, empty, error and success states must reflect the actual operation, including failures.
  • Keep technical and product documentation aligned when behaviour changes. Do not rewrite unrelated documents or broaden the product scope.

Verify the outcome

Select checks from requirements and risk. Use focused behaviour/regression tests, relevant lint/type/build commands, and the actual browser/device journey when available. For cross-layer work, confirm the result at the other end rather than only mocking every boundary.

Inspect test setup before execution: verify that destructive fixtures/reset commands target disposable data. Never point a test reset at a shared or production database. Use existing fake adapters for external side effects where possible.

Fix issues caused by the change, rerun the affected checks, and report unrelated failures separately. Do not remove assertions or update expected results just to match accidental implementation behaviour. Record checks that could not run.

Finish

Summarize delivered behaviour, changed surfaces, verification evidence and remaining limitations. Remove temporary diagnostics introduced by this work. A passing build does not establish product, design, security or performance acceptance; make claims only within the scope verified.

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.