neuronlayer 0.1.6 → 0.1.7

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/README.md +19 -26
  2. package/package.json +1 -1
package/README.md CHANGED
@@ -270,32 +270,25 @@ Each project has isolated data:
270
270
 
271
271
  ---
272
272
 
273
- ## Real-World Performance
274
-
275
- Benchmarked on Express.js (141 files, 21,487 lines of code):
276
-
277
- | Operation | WITHOUT MCP (grep) | WITH MCP (NeuronLayer) |
278
- |-----------|-------------------|------------------------|
279
- | Initial Setup | 0ms | ~28s (one-time indexing) |
280
- | Text Search | 56-60ms | ~10-50ms (cached) |
281
- | File Walk | 11ms | ~5ms (indexed) |
282
-
283
- **Honest Assessment:**
284
- - **grep wins** for simple text matching (56ms vs ~30ms)
285
- - **NeuronLayer wins** for semantic understanding and persistent memory
286
- - The ~28 second indexing is a one-time cost per session
287
- - After indexing, queries use cached embeddings
288
-
289
- **When NeuronLayer is Worth It:**
290
- - Long coding sessions (memory persists across context)
291
- - Complex queries ("how does auth work here?")
292
- - Architectural decisions (tracked & searchable)
293
- - Pattern consistency (learns your conventions)
294
- - Test awareness (knows what tests cover what)
295
-
296
- **When grep is Better:**
297
- - Quick one-off text searches
298
- - Small projects where indexing overhead isn't worth it
273
+ ## NeuronLayer vs grep
274
+
275
+ NeuronLayer **doesn't compete with grep** - Claude already has grep. They serve different purposes:
276
+
277
+ | grep does | NeuronLayer does |
278
+ |-----------|------------------|
279
+ | Text matching | **Semantic search** ("how does auth work here?") |
280
+ | Regex patterns | **Persistent memory** (decisions survive across sessions) |
281
+ | Fast file search | **Architectural awareness** (knows module relationships) |
282
+ | | **Pattern learning** (learns your coding conventions) |
283
+ | | **Context preservation** (survives conversation resets) |
284
+
285
+ **Real-world indexing** (Express.js - 141 files, 21k LOC):
286
+ - First-time indexing: ~28 seconds (one-time for new codebases)
287
+ - Subsequent sessions: Only changed files re-indexed (content hash check)
288
+ - Index persists in SQLite between sessions
289
+
290
+ **Use grep for:** "find all files containing `router`"
291
+ **Use NeuronLayer for:** "how does the routing system work?" or "what decisions were made about authentication?"
299
292
 
300
293
  ---
301
294
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "neuronlayer",
3
- "version": "0.1.6",
3
+ "version": "0.1.7",
4
4
  "description": "Persistent memory layer for AI coding assistants - MCP server that makes AI truly understand your codebase",
5
5
  "type": "module",
6
6
  "main": "dist/index.js",