agentleFS
ElasticKnown publisherelastic/terraform-provider-ec/.agents/skills/requirements-verification/SKILL.md

requirements-verification

Skill183 starsChanged 21 days agoApache-2.0View on GitHub

Analyze one OpenSpec capability spec (or a change delta) for internal consistency, implementation compliance, and test opportunities. Use when reviewing a single spec against its Go package. Do not use to gate an entire OpenSpec change before archive (that is openspec-verify-change, Phase 1.5). When a shell is available, run openspec validate first. Do not run acceptance tests (TF_ACC).

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    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:

    If validation fails, report CLI errors first and fix (or stop) before sections 1–3.

    line 29

≈ “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

Requirements verification

Analyze an OpenSpec spec (openspec/specs/<capability>/spec.md or a delta under openspec/changes/**/specs/) and produce three outputs:

  1. Internal consistency — requirements vs each other and vs the schema.
  2. Implementation compliance — whether the Go code meets each requirement.
  3. Test opportunities — unit tests the agent can run; acceptance tests named for a human/Buildkite only.

Input

  • Spec: path or entity name. Resolve to spec.md.
  • Implementation: from the spec’s Resource implementation: / Data source implementation: line. If that line is missing, treat it as incomplete (flag it) and locate the package via Metadata() / ec/provider.go.

Workflow

0. Structural validation (when a shell is available)

make check-openspec or OPENSPEC_TELEMETRY=0 ./node_modules/.bin/openspec validate --all

Single spec (optional): OPENSPEC_TELEMETRY=0 ./node_modules/.bin/openspec validate <capability> --type spec (e.g. deployment-traffic-filter).

This checks structure (Requirement/Scenario shape, SHALL/MUST in bodies) — not semantics or code compliance.

If validation fails, report CLI errors first and fix (or stop) before sections 1–3. If there is no terminal, note that validate was not run.

A file under this skill’s testdata/ is not part of openspec/ and is not seen by make check-openspec. Use it only when asked to exercise this skill on an intentionally incomplete spec.

1. Parse the spec

Extract: title/type name, implementation path, Purpose, Schema HCL if present.

List every ### Requirement: section by heading text (e.g. Ruleset identity). Do not invent REQ-xxx ids unless the heading already has them. Infer a category from content (see reference.md).

Flag as incomplete before going deep:

  • Missing ## Purpose
  • Missing Resource implementation: / Data source implementation:
  • A requirement body with no SHALL / MUST
  • A requirement with no #### Scenario:
  • Schema and requirements that disagree on required/optional/computed

2. Internal consistency

Apply reference.md (Consistency checks). Result: Consistent or a list of heading-vs-heading problems.

3. Implementation compliance

Resolve the package. Read schema.go, CRUD files, validators, and Schema.Version / UpgradeState if present. Mapping is not always expanders.go / flatteners.go: search for the functions that read API responses into Terraform models (e.g. ReadDeployment, ReadElasticsearches) and those that build API payloads (e.g. ElasticsearchPayload). For each requirement, mark Met / Not met / Unclear with evidence (file/function or “not found”). Mapping: reference.md.

4. Test opportunities

  • Unit: *_test.go in the entity package (and helpers). These the agent may run via make unit (no TF_ACC).
  • Acceptance: ec/acc/ (TestAcc…). List gaps as human/Buildkite suggestions. Never run make testacc or set TF_ACC=1.

5. Report

# Requirements analysis: <entity name>

**Document**: `openspec/specs/.../spec.md`
**Implementation**: `ec/ecresource/...` (or data source path)

## 1. Internal consistency

- **Result**: Consistent | Inconsistent
- **Inconsistencies** (if any): [Heading A] vs [Heading B]: ...
- **Incomplete** (if any): missing Purpose / implementation line / SHALL / scenarios / schema drift

## 2. Implementation compliance

| Requirement | Category | Status | Evidence |
| --- | --- | --- | --- |
| Ruleset identity | Identity | Met | create.go sets id from API |
...

**Summary**: X met, Y not met, Z unclear.

## 3. Test opportunities

| Requirement | Type | Suggested test | Verifies | Who runs |
| --- | --- | --- | --- | --- |
| Import by id | Unit | ImportState with id; state id set | Import passthrough | agent (`make unit`) |
| Create then read | Acceptance | `TestAccDeploymentTrafficFilter_basic` already covers create | Live create/read | human/Buildkite |

Reference

  • Authoring / CLI: dev-docs/high-level/openspec-requirements.md
  • Consistency checks and implementation mapping: reference.md
  • Code-path checklist: existing-entity-requirements/reference.md
  • Intentionally incomplete fixture (AC / skill exercise only): testdata/incomplete-spec.md

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.