Skip to content
AI & Agents
Skill

/api-design-reviewer

Use when reviewing a proposed REST or GraphQL API change before merge — checks contract clarity, backwards compatibility, errors, pagination, auth, and naming.

BOOST
From plugin
llm-agents-ecosystem-handbook
55510 skills
Install
$ npx -y skills add oxbshw/LLM-Agents-Ecosystem-Handbook --skill api-design-reviewer --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/api-design-reviewer

Context preview

The summary Claude sees to decide when to auto-load this skill.

Use when reviewing a proposed REST or GraphQL API change before merge — checks contract clarity, backwards compatibility, errors, pagination, auth, and naming.

SKILL.md

api-design-reviewer.SKILL.md
name: api-design-reviewer
description: Use when reviewing a proposed REST or GraphQL API change before merge — checks contract clarity, backwards compatibility, errors, pagination, auth, and naming.
version: 0.1.0
status: experimental
risk: low
tags: [engineering, read-only, writes-files]

API Design Reviewer

When to use

  • A new endpoint / mutation / schema is being proposed
  • A breaking change is on the table
  • An API draft needs a second pair of eyes before stakeholder review

When NOT to use

  • Pure implementation details (use `repo-auditor`)
  • UI / UX feedback
  • Cost / SLA tradeoffs (those need product discussion)

Inputs

| Name | Type | Required | Notes | |---|---|---|---| | `spec` | path or text | yes | OpenAPI / GraphQL SDL / Markdown design doc | | `mode` | `"public"` \| `"internal"` | no | Default `internal`. `public` raises the bar on backwards-compat. |

Outputs

A Markdown review with sections: Critical, Important, Nits, Open questions.

Workflow

1. Parse the spec; list the operations / types touched 2. For each operation, walk `references/api-review-checklist.md`:

  • Naming consistent with neighbors?
  • Resource model coherent?
  • Errors enumerated and typed?
  • Pagination present (where lists are returned)?
  • Auth + authz documented?
  • Idempotency on writes?
  • Versioning strategy?

3. For breaking changes, surface the contract delta explicitly 4. Group findings into **Critical** (blockers), **Important** (should fix), **Nits** (style) 5. Append **Open questions** for the author

References

  • [`references/api-review-checklist.md`](references/api-review-checklist.md)

Success criteria

  • 0 fabricated findings (every finding cites a section of the spec)
  • Critical/Important findings have a concrete suggested fix
  • Tone is collegial, not pedantic

Failure modes

  • Spec too large (> 200 ops) → split into modules and ask the author which to focus on
  • Internal jargon in the spec without definitions → list as Open questions, don't guess

Examples

  • Good: "Critical — `GET /users` returns an unbounded list. Add cursor pagination (suggested: `?cursor=...&limit=...`)."
  • Bad: "I think pagination might be useful here."
Read more
Ships withllm-agents-ecosystem-handbook

A practical operating manual for building, evaluating, securing, and shipping modern LLM agent systems. Modern agents are not "a prompt + a tool."

Get the whole plugin