devlensio — the DevLens analysis engine

The core engine behind DevLens. It turns a TypeScript / JavaScript / React / Next.js / Node.js repository into a typed code graph — nodes (components, hooks, functions, stores, routes, files, …) joined by typed edges — scores every node by architectural importance, optionally summarizes each node with an LLM (technical / business / security), and exposes a traversal/query API.

devlensio is a library + local server. The user-facing tools — the devlens CLI, the MCP server, the Agent Skill, and the Web UI — live in DevLens OSS and consume this package.

---

Install

npm install devlensio
# or: bun add devlensio

Requires Node 18+ (or Bun). An LLM provider key is needed only for summarization, not for structural analysis.

---

What it does (the pipeline)

Repo path
   │
[1] Fingerprint     detect language, framework, router, state manager, data layer, databases
[2] Filesystem scan extract routes (Next.js app/pages, Express, Fastify, Koa)
[3] Parse (ts-morph) walk every .ts/.tsx/.js/.jsx → nodes (typed params, return types, prop types)
[4] Edge detection  many detectors → CALLS, IMPORTS, READS_FROM, WRITES_TO, PROP_PASS, EMITS,
                    LISTENS, WRAPPED_BY, GUARDS, HANDLES, TESTS, USES
[5] Scoring         multi-pass importance scoring + noise filtering (no AI)
[6] Clustering      cohesive cluster assignment
[7] Summarize       (optional) topologically-ordered LLM summaries, checkpoint/resume, MapReduce
   │
   ▼
Graph persisted to ~/.devlens  →  queried via the traversal API / CLI / MCP / UI

Structural analysis is fast and deterministic; summarization is the only step that calls an LLM and reuses unchanged nodes across commits.

---

Public API

import {
  analyzePipeline,        // build the graph (nodes, edges, scores)
  runSummarization,       // generate technical/business/security summaries
  computeClusters,        // cohesive clustering
  buildGraphIndex,        // index nodes+edges for traversal
  getBlastRadius,         // upstream dependents ("what breaks if I change this")
  getKHop,                // downstream dependencies ("what this needs")
  getSubgraph,            // cohesive cluster around a seed
  findCycles,             // circular-dependency groups
  resolveConfig, initConfig,   // LLM provider config (~/.devlens/config.json)
  storage, queue,         // file-based graph storage + job queue singletons
} from "devlensio";

// Analyze a repo → graph
const result = await analyzePipeline("/path/to/repo", /* isGithubRepo */ false);
// result.allNodes, result.allEdges, result.nodeScores

// Query the graph
const index  = buildGraphIndex(result.allNodes, result.allEdges);
const impact = getBlastRadius(index, "src/auth/login.ts::login", { radius: 2 });
const cycles = findCycles(result.allNodes, result.allEdges);

Also exported: all core types (CodeNode, CodeEdge, NodeType, EdgeType, …), config helpers (maskConfig, writeConfig), pre-scan helpers (readPackageDependencies, categorizeLibrary), and EDGE_LABELS. See dist/index.d.ts for the full surface.

Node & edge types

• Node types: COMPONENT, HOOK, FUNCTION, STATE_STORE, UTILITY, FILE, ROUTE, TEST, STORY, THIRD_PARTY (+ internal GHOST).
• Edge types: CALLS, IMPORTS, READS_FROM, WRITES_TO, PROP_PASS, EMITS, LISTENS, WRAPPED_BY, GUARDS, HANDLES, TESTS, USES.
• Each node carries an importance score and (after summarization) a technical summary, a business summary, and a security assessment (none|low|medium|high + notes).

---

Configuration

Summarization config lives in ~/.devlens/config.json (set via initConfig/writeConfig, or env vars loaded with dotenv). Supported providers: Anthropic, OpenAI, OpenRouter, Gemini, Ollama (local).

LLM_PROVIDER=openrouter        # ollama | openai | anthropic | openrouter | gemini
LLM_MODEL=grok-4.1-fast
LLM_API_KEY=your_api_key       # not needed for ollama
LLM_BASE_URL=                  # e.g. http://localhost:11434 for Ollama

Graphs and config are stored under ~/.devlens and shared with all DevLens tools.

---

Repo layout

src/
├── fingerprint/   # detect framework, language, router, state, data layer, databases
├── filesystem/    # route detection (Next.js app/pages, Express, Fastify, Koa)
├── parser/        # ts-morph AST extraction → nodes
├── graph/         # edge detectors, traversal API, third-party libs, lookup maps
├── scoring/       # multi-pass importance scoring + noise filtering
├── clustering/    # cohesive cluster computation
├── summarizer/    # LLM summarization (technical/business/security), prompts, checkpoints
├── pipeline/      # analyzePipeline — orchestrates the whole analysis
├── jobs/          # job queue, concurrency, SSE progress events
├── storage/       # file-based graph persistence (~/.devlens)
├── config/        # provider config resolution
├── server/        # HTTP API server (consumed by the DevLens Web UI)
└── debug/         # exportGraph and dev utilities

---

Scripts

Script	Does
bun run dev	watch-mode HTTP server (src/server/index.ts)
bun run start	run the HTTP server
bun run build	tsc --project tsconfig.build.json → dist/ (the published artifact)
bun test	run the test suite
bun run export-graph	dump a graph for debugging

---

Relationship to DevLens OSS

devlensio is published to npm and consumed by DevLens OSS, which provides the devlens CLI (@devlensio/cli), the MCP server, the /devlens Agent Skill, and the Web UI on top of this engine. The CLI binaries bundle whatever version of devlensio resolves at build time, so engine fixes ship to users after a devlensio release and a bump of the dependency pin in DevLens OSS.

---

License

GNU Affero General Public License v3.0. If you run a modified version as a hosted service, you must release your modifications under the same license.