nodebench-mcp 2.0.0 → 2.1.0

package/README.md

@@ -33,6 +33,16 @@ Every additional tool call produces a concrete artifact — an issue found, a ri
 
 ---
 
+## Who's Using It
+
+**Vision engineer** — Built agentic vision analysis using GPT 5.2 with Set-of-Mark (SoM) for boundary boxing, similar to Google Gemini 3 Flash's agentic code execution approach. Uses NodeBench's verification pipeline to validate detection accuracy across screenshot variants before shipping model changes.
+
+**QA engineer** — Transitioned a manual QA workflow website into an AI agent-driven app for a pet care messaging platform. Uses NodeBench's quality gates, verification cycles, and eval runs to ensure the AI agent handles edge cases that manual QA caught but bare AI agents miss.
+
+Both found different subsets of the 75 tools useful — which is why v2.1 ships with `--preset` gating to load only what you need.
+
+---
+
 ## How It Works — 3 Real Examples
 
 ### Example 1: Bug fix
@@ -67,8 +77,11 @@ Tasks 1-3 start with zero prior knowledge. By task 9, the agent finds 2+ relevan
 ### Install (30 seconds)
 
 ```bash
-# Claude Code CLI (recommended)
+# Claude Code CLI — all 75 tools
 claude mcp add nodebench -- npx -y nodebench-mcp
+
+# Or start lean — 30 tools, ~60% less token overhead
+claude mcp add nodebench -- npx -y nodebench-mcp --preset lite
 ```
 
 Or add to `~/.claude/settings.json` or `.claude.json`:
@@ -187,6 +200,73 @@ Based on Anthropic's ["Building a C Compiler with Parallel Claudes"](https://www
 
 ---
 
+## Toolset Gating (v2.1)
+
+75 tools means ~19K tokens of schema per API call. If you only need core methodology, gate the toolset:
+
+### Presets
+
+```bash
+# Lite — 30 tools (verification, eval, gates, learning, recon)
+claude mcp add nodebench -- npx -y nodebench-mcp --preset lite
+
+# Core — 50 tools (adds flywheel, bootstrap, self-eval)
+claude mcp add nodebench -- npx -y nodebench-mcp --preset core
+
+# Full — all 75 tools (default)
+claude mcp add nodebench -- npx -y nodebench-mcp
+```
+
+Or in config:
+
+```json
+{
+  "mcpServers": {
+    "nodebench": {
+      "command": "npx",
+      "args": ["-y", "nodebench-mcp", "--preset", "core"]
+    }
+  }
+}
+```
+
+### Fine-grained control
+
+```bash
+# Include only specific toolsets
+npx nodebench-mcp --toolsets verification,eval,recon
+
+# Exclude heavy optional-dep toolsets
+npx nodebench-mcp --exclude vision,ui_capture,parallel
+
+# See all toolsets and presets
+npx nodebench-mcp --help
+```
+
+### Available toolsets
+
+| Toolset | Tools | What it covers |
+|---|---|---|
+| verification | 8 | Cycles, gaps, triple-verify, status |
+| eval | 5 | Eval runs, results, comparison |
+| quality_gate | 4 | Gates, presets, history |
+| learning | 4 | Knowledge, search, record |
+| recon | 5 | Research, findings, framework checks |
+| flywheel | 4 | Mandatory flywheel, promote, investigate |
+| bootstrap | 4 | Project setup, agents.md, self-implement |
+| self_eval | 6 | Trajectory analysis, health reports |
+| parallel | 10 | Task locks, roles, context budget, oracle |
+| vision | 3 | Screenshot analysis, UI capture |
+| ui_capture | 3 | Playwright-based capture |
+| web | 2 | Web search, URL fetch |
+| github | 3 | Repo search, analysis |
+| docs | 3 | Documentation generation |
+| local_file | 3 | CSV, XLSX, PDF parsing |
+
+`findTools` and `getMethodology` are always available regardless of gating — agents can discover tools on demand.
+
+---
+
 ## Build from Source
 
 ```bash

package/dist/__tests__/toolsetGatingEval.test.d.ts

@@ -0,0 +1 @@
+export {};