agentleFS
QuirkybrainQuirkybrain/claude-code-swe-framework/.claude/skills/integration-gate/SKILL.md

integration-gate

Skill1 starsChanged 53 days agoMITView on GitHub

Verify that individually STABLE Components remain correct when combined through Contract validation, integration, integration tests, and regression. Use before promoting any combined unit as stable.

2 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

7 rules. About 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run when two or more Components are ready to combine, when Contracts change, or when an integrated unit needs fresh evidence.

    line 14
  2. Very specific:

    Every participating Component is STABLE for the supplied revision.

    line 24
  3. Distinctive:

    Run Contract tests, integration tests, and relevant regression tests.

    line 36
  4. Very specific:

    Never integrate UNVERIFIED or FAILED inputs into the stable baseline.

    line 44
  5. Distinctive:

    Do not silently modify Contracts to make Components fit.

    line 46
  6. Distinctive:

    Use Failure Diagnosis for ambiguous failures and Failure Escalation after repeated local attempts.

    line 66
  7. Very specific:

    Stop at a recorded Gate verdict.

    line 70

≈ “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

Integration Gate

Purpose

Enforce that A = STABLE and B = STABLE does not imply A+B = STABLE.

Trigger

Run when two or more Components are ready to combine, when Contracts change, or when an integrated unit needs fresh evidence.

Inputs

  • Exact Component revisions and their Component Quality Gate evidence.
  • Versioned consumer/provider Contracts, integration design, and integration points.
  • Integration workspace/base, test strategy, environment/data, risks, and rollback plan.

Prerequisites

  • Every participating Component is STABLE for the supplied revision.
  • Contracts are sufficiently valid and compatible for attempted integration.
  • Integration workspace and ownership are explicit; unverified changes are excluded.

Procedure

  1. Perform the Stable Component Check and reject stale/mismatched evidence.
  2. Validate consumer/provider Contract versions, schemas, protocols, errors, and compatibility.
  3. Confirm workspace base, constituent revisions, rollback point, and absence of overlapping untracked integration work.
  4. Mark the combined unit UNVERIFIED.
  5. Integrate the smallest dependency-ready set in controlled order.
  6. Build/initialize the integrated unit.
  7. Run Contract tests, integration tests, and relevant regression tests.
  8. Validate cross-boundary data, error, security, performance, and lifecycle behavior as risk requires.
  9. Diagnose failures at interface, Contract, data, integration implementation, environment, design, architecture, or Requirement level.
  10. Record constituent revisions, commands, results, findings, and verdict.
  11. Mark the combined unit STABLE only on Gate PASS.

Decision Rules

  • Never integrate UNVERIFIED or FAILED inputs into the stable baseline.
  • Individual Components may remain STABLE after failure only when evidence isolates the defect to integration and their basis did not change.
  • Do not silently modify Contracts to make Components fit.
  • Regression depth follows actual combined impact.
  • Worktree isolation prevents file collision but does not merge, validate, or synchronize State automatically.

Outputs

  • Integration Gate Record: PASS, FAIL, BLOCKED, or USER_DECISION_REQUIRED.
  • Integrated revision and constituent/Contract revision manifest.
  • Contract, integration, regression, and cross-cutting evidence.
  • Integrated-unit state, failure classification, and next action.

Failure Cases

  • A Component is not truly STABLE or evidence is stale.
  • Contracts are incompatible or missing.
  • Integration environment/base is wrong or incomplete.
  • Cross-Component failure cannot be isolated.

Escalation

Use Failure Diagnosis for ambiguous failures and Failure Escalation after repeated local attempts. Use Change Impact when Contract/design changes invalidate Components. Request Clarification for product or authority choices.

Stop Conditions

Stop at a recorded Gate verdict. Proceed to the next integration increment only after the current combined unit is STABLE; never proceed automatically to release.

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.