agentleFS
neeraj181-dotneeraj181-dot/ClaudeSkills/.claude/skills/feature-integrity/SKILL.md

feature-integrity

Skill0 starsChanged 3 months agoApache-2.0View on GitHub

Determine whether a feature is actually complete end-to-end by tracing it through the full stack — UI, interaction, state, API, business logic, database/external services, response handling, and error states. Identifies broken links in the chain where a feature appears to exist but does not actually work. Never marks a feature complete based only on source code presence. Use when the user wants to verify a feature is genuinely working, or wants to know if a partially-built feature is ready for users.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before marking a ticket/story as complete.

    line 27
  2. Distinctive:

    After a rapid build session to verify nothing was missed.

    line 28

≈ “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

Feature Integrity

Creator: Neeraj Project: ClaudeSkills

Determine whether a feature is actually complete end-to-end — not just partially wired up.

Core principle: a button existing is not a feature working. A feature is only complete when the full chain works: UI → interaction → state → API → business logic → data → response → UI update → error handling. A break at any link means the feature is incomplete — even if 9 out of 10 links are solid.

This skill is not a QA test (qa-engineer), not a code audit (code-auditor), and not a completeness checker for documentation. It traces a specific feature through the full implementation chain and reports where the chain breaks.


When to activate

Activate when:

  • The user says "is this feature done?" or "is this ready?"
  • A feature was partially built and the user wants to know what's missing.
  • The user says "does the export actually work?" or "can users really do X?"
  • Before marking a ticket/story as complete.
  • After a rapid build session to verify nothing was missed.

Phase 1 — Identify the feature

Clearly define what feature is being checked:

  • What is the feature name?
  • What is the user-facing entry point? (button, link, form, route)
  • What is the expected end-to-end behavior?
  • What data does it read? What data does it write?

Phase 2 — Trace the full chain

For the feature, trace every link in this chain:

UI Element (button, link, form)
  ↓
Click/Submit Handler
  ↓
State Update (local or global)
  ↓
API Call (if any)
  ↓
Route/Controller
  ↓
Business Logic / Service Layer
  ↓
Database / External Service
  ↓
Response Construction
  ↓
State Update (response)
  ↓
UI Update (display result)
  ↓
Error Handling (all failure paths)

For each link, verify:

  • ✅ Exists — the code for this link is present.
  • ✅ Connected — the link is properly wired to adjacent links.
  • ✅ Functional — the logic does what it's supposed to.
  • ❌ Missing — the code does not exist yet.
  • ⚠️ Broken — the code exists but is not connected or not working.

Phase 3 — Check error paths

For every link, check what happens on failure:

  • Network error — does the UI show an error?
  • Invalid data — does validation catch it?
  • Unauthorized — does the user see a clear message?
  • Timeout — does the user know what happened?
  • Server error — is there a fallback or retry?

A feature that works on the happy path but crashes on errors is not complete.

Phase 4 — Check edge cases

For the feature, verify:

  • Empty state — what does the user see when there's no data?
  • Loading state — what does the user see while waiting?
  • Success state — is feedback clear and helpful?
  • Boundary conditions — max length, zero items, single item, many items.

Phase 5 — Classify completeness

For the overall feature:

FULLY COMPLETE
  All links exist, connected, and functional.
  Error paths handled. Edge cases covered.

MOSTLY COMPLETE
  Core chain works but minor gaps remain.
  One or two links missing or weak.

PARTIALLY COMPLETE
  Core chain has breaks.
  Feature cannot be used end-to-end.

BARELY STARTED
  Only UI or only backend exists.
  No connected chain yet.

NOT STARTED
  No code for this feature exists.

Phase 6 — Report broken links

For each broken or missing link:

LinkStatusFileIssueImpact
API Call❌ Missing—No API route for this actionFeature cannot save data
Error Handling⚠️ Brokenroute.tsCatches error but shows generic messageUser confused on failure

What this skill does NOT do

  • Does not test the running application (that is qa-engineer).
  • Does not audit all code quality (that is code-auditor).
  • Does not find bugs in existing features (that is incident-debugger).
  • Does not check performance (that is performance-optimizer).

This skill answers one question: "Is this specific feature actually complete from click to completion?"


Final output

# Feature Integrity Report
# Feature: [name]
# Chain Trace
# Links Verified
# Broken Links
# Edge Cases
# Completeness Classification
# What Remains

Guardrails: trace through actual code, not assumptions; verify each link exists AND is connected; check error paths, not just happy paths; never mark a feature complete based on source code presence alone; distinguish "code exists" from "code works"; keep the focus on one feature at a time; and never claim completeness without verifying the full chain.

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.