<div align="center">
<img src="assets/logo_1.png" alt="DevLens Logo" width="120" />
DevLens
Intelligent codebase visualizer.
Turn any JavaScript, TypeScript, React, Next.js, or Node.js repo into a living, queryable map — with functional summaries, technical summaries, and security analysis on every node.
  
Join the DevLens Cloud Waitlist →
</div>
---

<p align="center"><em>Click the image to watch the demo</em></p>
---
Table of Contents
- About
- Who is this for
- Quick Start
- The Problem DevLens Solves
- Benchmarks
- Ways to Use DevLens
- Web UI
- CLI
- Agent Skill
- MCP Server
- Configuration
- What DevLens Understands
- Repository Layout
- DevLens Cloud
---
About
DevLens is a codebase visualizer for JavaScript, TypeScript, React, Next.js, and Node.js projects. It builds a typed dependency graph of every component, hook, function, route, and store — with AI-powered summaries on each node — so you can explore, understand, and analyze your architecture in seconds instead of hours.
What makes it different? Instead of reading files one at a time, you (or your AI agent) query the pre-built graph. A node summary costs ~50 tokens vs ~2,000 tokens per file.
---
Who Is This For
Developers & Teams
- Onboard new devs in hours, not weeks — explore the graph instead of spelunking files.
- Review PRs with full context — see exactly what depends on each change.
- Run impact analysis before refactoring — "what breaks if I change this?"
- Catch circular deps, god-files, and coupling hotspots automatically.
- Keep living documentation — summaries stay fresh as code changes.
Engineering Leaders
- Get a bird's-eye view of your entire codebase in seconds.
- Spot architectural debt before it becomes a crisis.
- Understand what your team has been building — even across repos.
Students & Learners
- See how real codebases are designed — layers, patterns, data flow.
- Understand why things are connected, not just what each file does.
- Learn architecture patterns from production-grade projects.
AI-Augmented Developers
Your agent burns tokens re-reading files it's seen before. DevLens gives it a graph to query instead.
Coming soon — DevLens Cloud: Shareable graphs your whole team can access, cross-repo navigation, and giving graphical context to your AI agents for smarter code review and analysis — all without running anything locally.
---
Quick Start
1. Install
npm install -g @devlensio/cli
No Node.js? Use the standalone binary installer (no dependencies): Linux / macOS: ``
bash curl -fsSL https://raw.githubusercontent.com/devlensio/devlensOSS/main/scripts/install.sh | sh`Windows (PowerShell):`powershell irm https://raw.githubusercontent.com/devlensio/devlensOSS/main/scripts/install.ps1 | iex``
2. Init
cd your-project
devlens init
This sets up your LLM provider for AI summaries. Don't need AI? Skip this — structural analysis works offline.
3. Analyze
devlens analyze . --summarize
This builds a typed dependency graph of every component, hook, function, route, and store — with AI-powered summaries on each node.
4. Explore
devlens overview # big picture: framework, stats, central nodes
devlens find-nodes -t ROUTE # find every route in the app
devlens security # see security flags across the codebase
---
The Problem DevLens Solves
AI coding tools let you ship faster than ever — but that speed creates AI debt: code merged without understanding, agents re-discovering connections every session, new hires drowning in unfamiliar structure.
DevLens fixes this by pre-building a typed dependency graph of your entire codebase. Every node gets:
- Functional summary — what business purpose does this serve?
- Technical summary — how does it work?
- Security assessment — severity + explanation
Armed with this graph, you (or your AI agent) can understand the full architecture in ~50 tokens per node instead of ~2,000 per file.
---
Benchmarks
Tested across real-world tasks — architecture understanding, feature implementation, and bug finding — comparing the same model (DeepSeek V4 Flash, GLM 5.2, Kimi K2.6, Qwen 3.6) with and without DevLens.
Architecture Understanding (4 models, full DevLens MCP)
<br /><br />
<div align="center"> <img src="assets/01_arch_metrics.png" alt="Architecture benchmark — cost, tokens, steps comparison" width="90%" /> </div>
<br /><br />
| Metric | Without DevLens | With DevLens | Improvement | |--------|:--------------:|:------------:|:-----------:| | Avg cost per query | $0.163 | $0.075 | 54% cheaper | | Avg input tokens | 88,980 | 35,035 | 61% less | | Avg output tokens | 9,549 | 3,233 | 66% less | | Avg tool steps | 14.3 | 7.8 | 45% faster | | Structured architecture output | 50% | 100% | 2× more reliable | | Architectural debt discovered | 0% | 50% | Now discoverable | | Avg cache hit rate | 75.2% | 83.7% | +8.5pp |
<br /><br />
<div align="center"> <img src="assets/03_arch_savings_per_model.png" alt="Per-model savings" width="70%" /> <br /><br /> <img src="assets/02_arch_cache.png" alt="Cache hit rate comparison" width="70%" /> </div>
<br /><br />
Notable: Even the strongest model (DeepSeek V4 Flash) was 81% cheaper ($0.0035 vs $0.0185) and used 83% fewer input tokens with DevLens.
Feature Implementation & Bug Finding (DeepSeek V4 Flash)
5 prompts across implementation and debugging tasks — DevLens graph context only (no per-node summaries).
<br /><br />
<div align="center"> <img src="assets/04_prompt_metrics.png" alt="Prompt benchmark metrics" width="80%" /> </div>
<br /><br />
| Task | Input tokens saved | Cache improvement | |------|:-----------------:|:-----------------:| | Continue Watching feature | 24% less input (56.8k vs 74.9k) | +5.2pp cache | | Rate Limiting feature | 22% less input (32.1k vs 41.1k) | +8.3pp cache | | Error Handling audit | (comparable) | Comparable | | Profile Bug trace | 32% less input (36.9k vs 54.3k) | Similar |
Quality Impact (Architecture Task)
<br /><br />
<div align="center"> <img src="assets/09_quality_comparison.png" alt="Quality comparison" width="80%" /> <br /><br /> <img src="assets/08_quality_matrix.png" alt="Quality matrix" width="80%" /> </div>
<br /><br />
When asked to explain a codebase's architecture:
| Capability | Without DevLens | With DevLens | |-----------|:--------------:|:------------:| | Produced structured output | 50% | 100% | | Referenced specific graph metrics | 0% | 100% | | Identified architectural debt | 0% | 50% | | Named specific important files | 75% | 100% |
---
Ways to Use DevLens
Pick the interface that fits your workflow.
<img src="assets/web-icon.svg" width="20" style="vertical-align: middle" /> Web UI — Visual Exploration
For when you want to see your codebase laid out as an interactive graph.
Open the Web UI, paste your repo path, and explore a force-directed canvas — click any node to see its summaries, callers, callees, and security flags. Search, filter, diff commits across versions.
git clone https://github.com/devlensio/devlensOSS.git
cd devlensOSS && bun install && bun run dev
<img src="assets/cli-icon.svg" width="20" style="vertical-align: middle" /> CLI (@devlensio/cli) — Terminal Power
For scripts, CI, and when you want answers fast without leaving the terminal.
npm install -g @devlensio/cli
| Command | Description | Example | |---------|-------------|---------| | devlens analyze . | Analyze a repository into a DevLens graph | devlens analyze ./my-app --summarize | | devlens overview | Big picture — framework, stats, central nodes | devlens overview → "Next.js 15, 342 nodes, 12 routes, top node: Layout (9.2)" | | devlens blast-radius <nodeId> | What breaks if I change this? | devlens blast-radius "src/auth.ts::login" → "14 dependents affected" | | devlens cycles | Find circular dependencies | devlens cycles → Lists every circular import group | | devlens find-nodes -t <type> | Filter nodes by type (ROUTE, COMPONENT, etc.) | devlens find-nodes -t ROUTE → Lists all 12 routes | | devlens find-nodes --severity high | Find high-severity security issues | devlens find-nodes --severity high → "2 findings in auth module" | | devlens diff <from> <to> | Compare two analyzed commits | devlens diff abc123 def456 → "Added: AnalyticsTracker, Changed: CheckoutForm" | | devlens security | List every security issue | devlens security --min-severity high --json | | devlens config | View configuration with all saved providers | devlens config | | devlens config set | Interactive provider configuration with live model fetching | devlens config set | | devlens config --active <key> | Switch active provider without re-entering credentials | devlens config --active openai:deepseek | | devlens config --remove <key> | Remove a saved provider entry | devlens config --remove anthropic:anthropic |
Each command supports --json for piping into scripts and CI pipelines, -v/--verbose for diagnostics, and --quiet for minimal output.
Full reference:
src/cli/README.md— every command with examples and options.
<img src="assets/skill-icon.svg" width="20" style="vertical-align: middle" /> Agent Skill (@devlensio/skill) — AI-Powered Understanding
The most powerful way to use DevLens. Your AI agent normally reads files one at a time — the DevLens Skill teaches it to query the pre-built graph instead.
npx @devlensio/skill install
Then reload your tool and use /devlens in Claude Code, Cursor, or Kilo:
| Command | Description | Example | |---------|-------------|---------| | /devlens architecture | Full system overview — stack, modules, routes, patterns | /devlens architecture → Returns structured report: framework, 12 routes, 3 modules, security posture | | /devlens security-analysis | Prioritized security findings with exploit notes | /devlens security-analysis high → "SQL injection in loginUser (reach: 14 nodes)" | | /devlens impact <symbol> | Blast radius — what breaks if you change this? | /devlens impact loginUser → "14 dependents across 3 modules" | | /devlens tech-debt | Circular deps, coupling hotspots, god-files | /devlens tech-debt → "3 cycles, Navbar has 28 dependents" | | /devlens guard [target] | Warn before editing high-risk code | /devlens guard → "⚠️ authMiddleware affects 22 dependents" | | /devlens onboard | Generate ONBOARDING.md for new devs | /devlens onboard → Writes full onboarding doc to repo root | | /devlens explain [path] | Understand a module with callers/callees | /devlens explain src/api/ → Walks through all API handlers | | /devlens find <name> | Locate any component, hook, or function | /devlens find Button → "3 matches across components and tests" | | /devlens changes [range] | Explain what changed and its impact | /devlens changes yesterday → "3 files, 2 features, 1 bug fix" | | /devlens diagram [type] | Mermaid diagrams of architecture or flows | /devlens diagram architecture → Generates layered module diagram |
Full reference:
packages/skill-installer/README.md— all subcommands with examples, install options, and supported tools.
<img src="assets/mcp-icon.svg" width="20" style="vertical-align: middle" /> MCP Server — For Any MCP-Compatible AI Agent
Wire DevLens into any MCP client (Claude Desktop, IDE plugins, etc.). Bundled inside the CLI — exposes 14 tools over the Model Context Protocol.
devlens mcp # stdio mode
claude mcp add devlens -- devlens mcp # register in Claude Code
devlens mcp http -p 7000 # HTTP mode
Each MCP tool is a query into the pre-built graph — your agent can:
list_analyzed_repos— see what repos are already analyzedget_repo_overview— framework, stats, route count at a glancefind_nodes— search by name, type, file, severity, or scoreget_blast_radius— check impact before refactoringget_security_issues— rank all security findingslist_cycles— find circular dependenciesanalyze_changes— diff two analyzed commits
Full reference:
src/mcp/README.md— tool reference, examples, registration, and configuration.
---
Configuration
Config lives in ~/.devlens/config.json — set via devlens init or devlens config.
| Provider | Recommended model | Notes | | :-- | :-- | :-- | | Ollama (local) | qwen2.5-coder:7b | Free, local, 8GB+ RAM | | OpenAI | gpt-4o-mini | Fast, cost-effective | | Anthropic | claude-haiku-4-5 | Best cost/quality for summaries | | DeepSeek | deepseek-v4-flash | Strong code model | | OpenRouter | deepseek-v4-flash or mimo-v2.5 | Best cost/quality balance | | Gemini | gemini-2.0-flash | Fast, large context |
# Interactive setup — picks from a catalog and fetches live model lists
devlens config set
# Non-interactive scripting
devlens config --provider openai --provider-name deepseek --model deepseek-v4-flash --api-key <key>
# Switch between saved providers without re-entering credentials
devlens config --active openai:deepseek
Models are discovered dynamically from each provider's /models endpoint — no hardcoded model lists. Custom OpenAI- or Anthropic-compatible endpoints can be added through the interactive flow.
---
What DevLens Understands
Node types: COMPONENT, HOOK, FUNCTION, STATE_STORE, UTILITY, FILE, ROUTE, TEST, STORY, THIRD_PARTY
Route types: Next.js (app & pages), Express / Fastify / Koa, React Router / TanStack Router / wouter
Edge types: CALLS, IMPORTS, READS_FROM, WRITES_TO, PROP_PASS, EMITS, LISTENS, WRAPPED_BY, GUARDS, HANDLES, TESTS, USES, NEXTJS_API_CALL, NAVIGATES_TO
Per node: Importance score + functional summary + technical summary + security assessment
---
Repository Layout
devlensOSS/
├── src/
│ ├── cli/ # `devlens` CLI (commander program + commands)
│ ├── core/ # Shared query core (CLI + MCP — never drift)
│ ├── mcp/ # MCP server (stdio + HTTP), 14 tools
│ └── server/ # Backend API for the Web UI
├── frontend/ # Next.js 15 graph visualizer (Cytoscape)
├── plugins/devlens/ # Agent Skill (Claude plugin source)
├── packages/skill-installer/ # @devlensio/skill — npx installer
├── bin/ # npm launcher
├── npm/<platform>/ # 5 prebuilt binary packages
├── scripts/ # Release tooling
└── server.json # MCP registry manifest
The analysis engine lives in the separate devlensio package.
---
DevLens Cloud
A hosted version is in development:
- Shareable graphs your whole team can access
- Cross-repo navigation — understand your entire org
- Give graphical context to AI agents for smarter code review and analysis
- No local setup required











