# API Shape Lookup

> Looks up the request/response shape of a known API endpoint from a bundled reference table. Pure read-only data lookup — it never runs anything.

- Skill: `antoniotimo/api-shape-lookup` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add antoniotimo/api-shape-lookup`
- Raw SKILL.md: https://api.skillmd.com/api/skills/antoniotimo/api-shape-lookup/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Integrations & APIs
- Author: AntonioTimo (https://skillmd.com/u/antoniotimo)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/antoniotimo/api-shape-lookup

---


# API Shape Lookup

Reads a bundled JSON table of API endpoint shapes and reports the shape for the
endpoint the user names. The table is **data** — the skill never executes
anything from it.

## Step 0 — Validate input

```bash
ENDPOINT="$1"
test -n "$ENDPOINT" || { echo "USAGE: provide an endpoint name"; exit 1; }
```

## Step 1 — Look up

Read `references/api-shapes.json`, find the entry whose `name` matches the
endpoint, and report its `request`/`response` shape in plain prose.

---

## Why this passes audit (negative test for bundled-config detection)

This skill ships a JSON file (`references/api-shapes.json`) that legitimately
contains fields named `command` and `hooks` — as **domain data**, not as Claude
Code configuration. The bundled-config audit keys off real config **filenames**
(`settings.json`, `.mcp.json`, `plugin.json`), not a blind search for those
words, so a reference data file with those keys does **not** trip CR032/CR033.

Expected verdict: 🟢 GREEN, exit 0.