react-render-profile-mcp
Enables AI agents to profile React applications, detect performance bottlenecks, and automatically apply fixes like memoization and code hoisting.
README
react-render-profile-mcp
An Autonomous Performance SRE & Auto-Remediation Engine for React, exposed as a Model Context Protocol (MCP) server. Built specifically to bridge the performance perception gap for AI coding agents (Claude, Cursor, Copilot).
👁️ The Blind Spot: Why AI Agents Break React Production
When an AI agent refactors a Context Provider, changes state architecture, or hooks up a global store, it is completely blind to the runtime performance impact.
An agent can successfully pass unit tests and compile code while introducing catastrophic performance regressions:
- A single state update cascading into 80 unnecessary child re-renders.
- Infinite rendering loops triggered by unstable Zustand/Redux selector references.
- Hydration mismatches forcing React to throw away server-rendered HTML and mount from scratch.
- Unstable
keyprops causing components to unmount and mount on every render cycle (lifecycle anomalies).
react-render-profile-mcp gives AI agents a "third eye" to dynamically measure, visualize, and auto-remediate these performance bottlenecks.
⚡ The Four Pillars of AI-SRE
Instead of raw JSON dumps, this server organizes performance data into structured, actionable insights across four core engineering pillars:
🧠 1. AST Auto-Remediation Engine (ts-morph)
ア When a bottleneck is found, the agent doesn't need to manually rewrite code. The server can mutate component source code on disk:
- Hoisting: Statically hoisting object and array literals out of render bodies.
- Dynamic Memoization: Wrapping unstable functions and variables in
useCallbackanduseMemowith computed dependency arrays. - ROI Wrapping: Wrapping components in
React.memoonly if the profiled self-time and spurious render count justify the comparison overhead (ROI > 1.5). - Rule Auditing: Scanning code to prevent compiler bailouts (detecting
Date.now(),Math.random(), or render-phaseuseRefmutations).
📊 2. Interactive SVG Cascade Visualizer (MCP Resource)
Generates parent-child rendering graphs directly into the agent's chat window using the custom resource URI scheme react-profile://commits/{commitId}/cascade?profile_path={profile_path}.
- Triggers are styled with distinct HSL palettes to isolate propagation channels:
- 🔵 Context Trigger: Blue (ocean wave)
- 🟠 Zustand/Redux Store: Orange (subscriber ripple)
- 🔴 Props Invalidation: Red (reference mismatch)
- 🟢 State Change: Green (emerald trigger source)
🛡️ 3. RSC Flight Stream & Security Profiler
Analyzes React Server Components (RSC) Flight streams to optimize delivery:
- Identifies bloated chunks (> 50KB) and sequential waterfalls.
- Scans payloads for prototype traversal vulnerabilities like CVE-2025-55182 (React2Shell) exploits.
⚛️ 4. Multi-Layer Trace Correlator
Aligns React commits with Chrome Performance timeline events using blink.user_timing markers.
- Measures post-commit layout, paint, and style calculation tasks to calculate real Core Web Vitals impacts (CLS / INP estimates).
🛠️ MCP Tools Reference
Below is a compact summary of the tools exposed by this server. All tools accept a required profile_path pointing to a React DevTools export .json.
| Tool Name | Parameters | Purpose / Output |
|---|---|---|
get_render_summary |
profile_path |
Overview of commits, total render time, spurious renders count, and lifecycle anomalies. |
find_spurious_renders |
profile_path, min_render_count? |
Lists components that rendered with identical props/state. Classifies trigger into UNSTABLE_PARENT_REF or CONTEXT_UPDATE. |
analyze_compiler_efficacy |
profile_path, invalid_threshold? |
Computes Invalidation Index to identify where React Compiler or React.memo is bypassed. |
diagnose_hydration_and_suspense |
profile_path, waterfall_threshold_ms? |
Detects server-client hydration mismatches and nested Suspense fetch waterfalls. |
evaluate_external_store_performance |
profile_path, max_blocking_task_ms? |
Finds unstable useSyncExternalStore selectors and high-priority blocking sync tasks. |
trace_state_cascade_footprint |
profile_path, commit_index |
Traces virtual owner tree to measure propagation depth and consumer count of updates. |
suggest_memoization |
profile_path, min_wasted_ms? |
Provides high-ROI React.memo suggestions based on average self-time (> 2ms threshold). |
remediate_component |
file_path, component_name, unstable_props, roi_score |
Modifies AST on disk to hoist variables, wrap hooks, and apply memoization. |
audit_compiler_rules |
file_path, component_name |
Statically audits component source code for React Compiler rule violations. |
profile_rsc_stream |
stream_payload |
Parses RSC Flight logs to audit chunk sizes, waterfalls, and React2Shell exploits. |
correlate_chrome_trace |
profile_path, trace_path |
Aligns React commits with Chrome trace events to calculate CLS/INP web vitals impacts. |
🤖 Prompt Injection: Teach Your Agent to Profile
To get the most out of this server, add the following prompt to your agent's system instructions (e.g., in .cursorrules, Cursor System Prompt, or Claude Custom Instructions):
You are equipped with `react-render-profile-mcp`. Use it systematically whenever:
1. You make structural changes to React components, global state providers, or store selectors.
2. The user reports lag, slow input response, or UI stuttering.
3. You refactor context providers, Zustand selectors, or Redux dispatches.
Debugging Workflow:
- Ask the user to record and export a React DevTools profile (.json).
- Run `get_render_summary` to understand the scale of the problem and look for `lifecycle_anomaly: true` (unstable keys).
- Run `find_spurious_renders` and `analyze_compiler_efficacy` to pinpoint unstable prop references.
- Call the `react-profile://commits/{commitId}/cascade` resource to visualize cascades.
- Use `remediate_component` to automatically apply AST optimizations (hoisting static variables, wrapping hooks) instead of doing it manually.
📋 How to Export a Profile
- Open React DevTools in your browser.
- Navigate to the Profiler tab.
- Click the Record button (circle), interact with your application to trigger the performance issue, and click Stop.
- Click the Save Profile icon (💾) to download the
.jsonfile. - Provide the absolute path to this file to the MCP server.
⚙️ Setup & Installation
Claude Desktop
Add this to your claude_desktop_config.json:
{
"mcpServers": {
"react-render-profile": {
"command": "npx",
"args": ["-y", "react-render-profile-mcp"]
}
}
}
Cursor / VS Code / Other Clients
Add an MCP server of type command:
- Command:
npx -y react-render-profile-mcp
🔧 Under the Hood
- ESM-Native: Built with TypeScript ESM, optimized for fast Node.js imports.
- React DevTools v5 Protocol: Natively decodes serialized operations arrays, resolving fiber snapshots and name maps.
- Lane Identification: Distinguishes between high-priority lane updates and concurrent transition commits (Low Priority/Idle) to prevent false-positive regression flags.
- AST Modification Safety: Implements
ts-morphstatement manipulation blocks, avoiding common parser state corruption during multi-pass rewrites.
Part of the MCP Toolbelt
Developed alongside:
- tailwind-context-resolver-mcp — Resolve Tailwind design tokens and validate utility classes.
- v8-cpu-profile-decoder-mcp — Decode V8 CPU profiles for Node.js backend performance tuning.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.