DevMemory is a long‑term memory for AI coding agents. Make AI knowledge searchable.
Project description
DevMemory 🧠🚀
AI code attribution that answers "why?" — for developers and AI agents.
Built on Git AI for capture and Redis Agent Memory Server for semantic search and recall.
For local mode we use sqlite
Status: currently in Beta.
AI code attribution that answers "why?" — for developers and AI agents.
DevMemory tracks which AI tool wrote which line of code, then makes that knowledge searchable. Ask "why did we use this pattern?" and get answers backed by commits, prompts, and context.
Quick Install
# One-line install (local mode)
curl -sSL https://raw.githubusercontent.com/AI-Provenance/ai-dev-memory/main/scripts/install-simple.sh | bash
Then in your project:
devmemory install --mode local
Choose Your Setup
Option 1: Local (SQLite) — Free, No Infrastructure
For tracking AI code attributions locally without external services.
# 1. Install
pip install devmemory
# 2. Set up in your repo
cd your-project
devmemory install --mode local
# 3. Start coding — commits auto-sync
git commit -m "feat: new login" # AI attribution tracked automatically
devmemory attribution lookup path/to/file.py # See who wrote what
What's included:
- AI code line attribution (SQLite)
- Git hooks for auto-sync
devmemory attributioncommands- Sentry error tracking with AI attributions
Sentry Integration
When an error hits Sentry, see which AI tool and model wrote the code that caused it.
Python:
import sentry_sdk
from devmemory.sentry import create_before_send
sentry_sdk.init(
dsn="YOUR_SENTRY_DSN",
before_send=create_before_send(),
)
Node.js / Next.js:
npm install @aiprovenance/sentry
import { createDevMemoryBeforeSend } from "@aiprovenance/sentry";
Sentry.init({
dsn: process.env.NEXT_PUBLIC_SENTRY_DSN,
beforeSend(event, hint) {
return createDevMemoryBeforeSend()(event, hint);
},
});
Auto-detects repoId from git remote or DEVMEMORY_REPO_ID env var.
This adds ai_model, ai_tool, author, commit_sha, and other tags to every Sentry event.
Option 2: Cloud (Redis AMS) — Full Features
Note: Full Cloud mode is coming soon. Some features below require additional setup.
For teams who want semantic search, team stats, and AI agent memory.
# 1. Install
pip install devmemory
# 2. Start Redis AMS
docker compose --profile debug up -d
# 3. Set up in your repo
cd your-project
devmemory install --mode cloud
# Start using
devmemory why src/auth.py # Ask why a file looks this way
devmemory search "how do we auth?" # Semantic search
devmemory stats # AI vs Human code stats
What's included:
- Everything in Local mode
- Semantic memory search
- Team code statistics
- Cursor/Claude agent integration
- Context briefings on branch switch
Sentry integration (Cloud): Error tracking with AI attribution is coming soon — requires API call to enrich events.
Why DevMemory?
| For Developers | For AI Agents |
|---|---|
| "Why did we choose this pattern?" | Start with repo context, not from scratch |
| "Who wrote this — AI or human?" | Remember what previous agents learned |
| Team AI vs Human code stats | Reuse knowledge across sessions |
With Sentry Integration: Errors automatically include AI attribution data — see exactly which AI tool and model generated the code that caused the crash.
Commands
Local Mode
| Command | Description |
|---|---|
devmemory attribution lookup <file> |
See who wrote each line |
devmemory attribution history <file> |
View attribution history |
devmemory sync |
Sync Git AI notes to SQLite |
devmemory status |
Check system health |
Cloud Mode
| Command | Description |
|---|---|
devmemory why <file> |
Explain why a file/function looks this way |
devmemory search <query> |
Semantic search across all memories |
devmemory stats |
AI vs Human code contribution |
devmemory attribution lookup <file> |
See who wrote each line |
devmemory attribution history <file> |
View attribution history |
devmemory sync |
Sync Git AI notes to AMS |
devmemory status |
Check system health |
Requirements
- Python 3.10+
- Git AI — for capturing AI code attribution
- (Cloud mode) Docker — for Redis AMS
Learn More
Stop re-explaining your code. Let DevMemory remember.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file devmemory-0.1.30.tar.gz.
File metadata
- Download URL: devmemory-0.1.30.tar.gz
- Upload date:
- Size: 9.7 MB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
70fbfff7bcec590c94f06f0ce7679936ce77a6a9cccc5abe57b6e141f94ee80d
|
|
| MD5 |
97709d0c6fd3727fca57548ec18ed5b3
|
|
| BLAKE2b-256 |
b2a76f82de4cd1fa94434293f52356caa33a66594b9a62824d89810889134d08
|
File details
Details for the file devmemory-0.1.30-py3-none-any.whl.
File metadata
- Download URL: devmemory-0.1.30-py3-none-any.whl
- Upload date:
- Size: 98.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
9209b8200994ac720225ebdbcfa6b52f9232cec3e5598324de78f3061384e367
|
|
| MD5 |
9530cc7dabb5b6148ead1eb3435bd709
|
|
| BLAKE2b-256 |
cafe8cb5742ffe65e739e93aadf57f2afece7e6482ffc6350b6e331af1d077e2
|