@glubean/cli 0.1.21 → 0.1.22

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@glubean/cli",
-  "version": "0.1.21",
+  "version": "0.1.22",
   "type": "module",
   "bin": {
     "glubean": "./bin/gb.js"
@@ -27,9 +27,9 @@
     "yaml": "^2.7.0",
     "tsx": "^4.19.0",
     "@glubean/sdk": "0.1.19",
+    "@glubean/runner": "0.1.19",
     "@glubean/redaction": "0.1.16",
-    "@glubean/scanner": "0.1.18",
-    "@glubean/runner": "0.1.19"
+    "@glubean/scanner": "0.1.18"
   },
   "devDependencies": {
     "@types/node": "^22.0.0",

package/templates/claude-skill-glubean-test.md

@@ -1,6 +1,6 @@
 ---
 name: glubean
-description: Generate, run, and fix Glubean API/browser tests. Uses cheatsheet docs for SDK patterns and CLI for execution.
+description: Generate, run, and fix Glubean API/browser tests. Uses MCP for AI-optimized execution (schema inference + truncation) and lens docs for SDK patterns.
 context: fork
 allowed-tools:
   - Read
@@ -9,6 +9,12 @@ allowed-tools:
   - Glob
   - Grep
   - Bash
+  - mcp__glubean__glubean_run_local_file
+  - mcp__glubean__glubean_discover_tests
+  - mcp__glubean__glubean_list_test_files
+  - mcp__glubean__glubean_diagnose_config
+  - mcp__glubean__glubean_get_last_run_summary
+  - mcp__glubean__glubean_get_local_events
 ---
 
 # Glubean Test Generator
@@ -19,42 +25,44 @@ You are a Glubean test expert. Generate, run, and fix tests using `@glubean/sdk`
 
 1. **Secrets → `.env.secrets`**, public vars → `.env`. NEVER inline as `const`.
 2. **Use `configure()`** for HTTP clients — never raw `fetch()`.
-3. **Tags on every test** — `["smoke"]`, `["api"]`, `["e2e"]`, etc.
-4. **Teardown** any test that creates resources.
-5. **IDs**: kebab-case, unique across project.
-6. **Type responses**: `.json<{ id: string }>()`, never `.json<any>()`.
-7. **One export per behavior**: each `export const` is one test case.
-8. **Imports**: `import { test, configure } from "@glubean/sdk"` — always the npm package name.
+3. **All values use `{{KEY}}`** for env references, bare strings for literals.
+4. **Tags on every test** — `["smoke"]`, `["api"]`, `["e2e"]`, etc.
+5. **Teardown** any test that creates resources.
+6. **IDs**: kebab-case, unique across project.
+7. **Type responses**: `.json<{ id: string }>()`, never `.json<any>()`.
+8. **One export per behavior**: each `export const` is one test case.
 
 ## Workflow
 
-1. **Load cheatsheet** — check `.glubean/docs/index.md` exists and is fresh.
+1. **Load lens docs** — check `.glubean/docs/index.md` exists and is fresh.
    - Missing → run `npx glubean docs pull` via Bash, then continue.
    - Stale (`.glubean/docs/.pulled_at` is older than 1 day) → run `npx glubean docs pull` to update.
-   - Read `.glubean/docs/index.md` to see all available patterns and SDK capabilities.
+   - Read `.glubean/docs/index.md` to see all available patterns, plugins, and SDK capabilities.
 
 2. **Read relevant patterns** — based on the user's request, read 1-3 pattern files from `.glubean/docs/patterns/`.
-   For example: `configure.md` + `crud.md` for a CRUD test, or `browser.md` for E2E.
+   For example: `configure.md` + `crud.md` for a CRUD test, or `auth.md` for API key setup.
    Also read `.glubean/docs/sdk-reference.md` if you need the full API surface.
 
-3. **Read the API spec** — check `context/*-endpoints/_index.md` (pre-split specs). If found, read the index
+3. **Explore the API** — use MCP tool `glubean_run_local_file` with `includeTraces: true` on an existing
+   test file (or a quick smoke test) to see response schemas. Each trace includes:
+   - `responseSchema` — inferred JSON Schema (field names, types, array sizes)
+   - `responseBody` — truncated preview (arrays capped at 3 items, strings at 80 chars)
+   Use `responseSchema` to understand the API structure before writing assertions.
+
+4. **Read the API spec** — check `context/*-endpoints/_index.md` (pre-split specs). If found, read the index
    and only open the specific endpoint file you need. If no split specs, search `context/` for OpenAPI specs
-   (`.json`, `.yaml`). If a spec is larger than 50K, suggest `npx glubean spec split context/<file>`. If
-   no spec found, ask the user for endpoint details.
+   (`.json`, `.yaml`). If no spec found, ask the user for endpoint details.
 
-4. **Read existing tests** — check `tests/`, `explore/`, and `config/` for patterns, configure files, and
+5. **Read existing tests** — check `tests/`, `explore/`, and `config/` for patterns, configure files, and
    naming conventions already in use. Follow the project's existing style.
 
-5. **Write tests** — generate test files following the patterns from the cheatsheet and the project's conventions.
+6. **Write tests** — generate test files following the patterns from the lens docs and the project's conventions.
 
-6. **Run tests** — execute via Bash:
-   ```bash
-   npx glubean run <file>           # Run single file
-   npx glubean run --filter <tag>   # Run by tag
-   npx glubean run --verbose        # Show traces and assertions
-   ```
+7. **Run tests** — prefer MCP, fall back to CLI:
+   - **MCP** (preferred): `glubean_run_local_file` — structured results with schema-enriched traces.
+   - **CLI** (fallback): `npx glubean run <file> --verbose`
 
-7. **Fix failures** — read the output, fix the test code, and rerun. Repeat until green.
+8. **Fix failures** — read the structured failure output, fix the test code, and rerun. Repeat until green.
 
 If $ARGUMENTS is provided, treat it as the target: an endpoint path, a tag, a file to test, or a natural
 language description.
@@ -69,7 +77,7 @@ data/            # Test data files (JSON, CSV, YAML)
 context/         # OpenAPI specs and reference docs
 .env             # Public variables (BASE_URL)
 .env.secrets     # Credentials — gitignored
-.glubean/docs/   # SDK cheatsheet (auto-pulled, gitignored)
+.glubean/docs/   # SDK lens docs (auto-pulled, gitignored)
 package.json     # Runtime config, dependencies
 ```