qiuyanxin

Give every AI agent persistent memory of your team's knowledge. No vector DB, no RAG — just Git + BM25 + 114 tokens per session.

11
0
100% credibility
Found Apr 09, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

sp-context is a command-line tool that manages a shared folder of team knowledge documents, enabling AI coding assistants to search and access persistent context without databases or complex setups.

How It Works

1
💡 Discover the helper

You hear about sp-context, a simple way to give your AI coding buddy instant access to your team's shared wisdom like decisions and tips.

2
📥 Get it ready

Download the tool to your computer and prepare it so you can use helpful commands anytime.

3
📁 Set up your team's notes

Create a special folder for your team's important info, like company details, plans, and lessons learned.

4
🔗 Link to your AI chat

Tell your AI helper to peek at the notes folder every time a new chat starts, so it always knows the big picture.

5
Save new knowledge

Quickly add notes about meetings, ideas, or updates with a simple command, keeping everything organized.

6
🔍 Search and ask away

Type a question to your AI, and it finds the right notes, shares answers with sources, and feels like a team expert.

🎉 Team smarts unlocked

Now your AI remembers your team's knowledge forever, saving time and making coding chats super productive.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is sp-context?

sp-context is a TypeScript CLI built on Bun that gives every AI agent—like Claude Code, Cursor, or Copilot—persistent memory of your team's knowledge via a plain Git repo. It injects a tiny ~100-token catalog overview at session start, then lets agents query with `sp search`, fetch full docs via `sp get`, or browse categories like decisions and playbooks using BM25 search. Ditch vector DBs and RAG pipelines; store architecture notes, lessons learned, and status updates in Git for versioned, shared access—perfect for giving GitHub Copilot context or read-only access to private repos.

Why is it gaining traction?

It skips heavy infra for local-first simplicity: `bun install`, init a repo, and hook into agent sessions—no vendor lock-in or costs. Progressive loading saves tokens (overview first, details on-demand), while Unix-style pipes output JSON for scripting, and built-in governance like `sp doctor` flags duplicates or stale docs. Teams love the Git workflow for easy collaboration, like giving GitHub collaborator read access without complex permissions.

Who should use this?

AI-heavy dev teams using Claude Code or Cursor for pair-programming, needing quick access to company context, decisions, or client playbooks. Solo devs or small squads tired of re-explaining architecture every session, especially those already in GitHub for private repo sharing with read-only links. Ideal for frontend/backend leads capturing meetings or status updates without RAG overhead.

Verdict

Try it if you're evaluating lightweight knowledge bases for AI agents—solid docs and CLI make onboarding fast, despite 11 stars and 1.0% credibility signaling early maturity. Pair with GitHub permissions for team sync; scale cautiously until more adoption hits.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.