obsfx

obsfx / promptscout

Public

Enrich coding agent prompts with codebase context using a local LLM

agent ai coding-agent context-engineering llamacpp
29
2
100% credibility
Found Feb 18, 2026 at 25 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A desktop tool that scans your codebase and attaches relevant files, code snippets, and history to make AI coding prompts much more effective, all using a local brain with no internet needed.

How It Works

1
🔍 Hear about PromptScout

You're chatting with an AI helper about your code, but it keeps asking for more details - then you find this handy tool that adds the right bits from your project automatically.

2
📥 Bring it home

Download and add the tool to your computer in a quick, simple step.

3
🧠 Wake up the helper

Run a one-time setup to grab the small thinking brain that lives on your machine and scans your code for free.

4
💬 Magic question time

Type a question like 'fix the login bug' and watch it instantly pull in the exact files, code spots, and changes from your project to supercharge your words.

5
Ready to share
📄
Paste it yourself

Stick the smarter version into any AI helper and see it understand your code perfectly.

🔌
Auto-boost in app

It hooks right into your AI coding friend, adding smarts behind the scenes every time.

6
📜 Relive the wins

Peek back at past questions and their boosted versions anytime to reuse or learn.

🎉 AI gets your code

Now your AI helper skips the guessing and dives straight into fixing, building, or explaining your project like it lives there.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 25 to 29 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 promptscout?

Promptscout is a TypeScript CLI that enriches coding agent prompts with relevant codebase context using a local LLM, like Qwen 3 4B. Feed it a prompt like "fix auth token refresh," and it scans your repo with ripgrep and git to append file paths, code snippets, definitions, and commit history—saving your paid AI agent's tokens on discovery. No cloud, no API keys; just run locally and paste the boosted prompt into Claude, Cursor, or Copilot.

Why is it gaining traction?

Unlike cloud-heavy enrichers (think carles domingo enrich github or snowplow enrich github), promptscout runs offline with GPU acceleration, delivering results in ~2s via Metal/CUDA. Its Claude Code plugin hooks prompts automatically, showing summaries like "+5 files, +3 sections," while falling back silently if needed. History tracking and dry-run mode make iteration painless for agent workflows.

Who should use this?

Backend devs debugging modules in monorepos, frontend teams tracing React hooks or auth flows, or anyone prompting agents on unfamiliar codebases. Ideal for TypeScript/JS projects where ripgrep shines, or Swift/Go/Rust repos needing quick context grabs before enriching mcp github-style pipelines.

Verdict

Grab it if you're deep in AI-assisted coding—early 18 stars and 1.0% credibility score mean it's raw, but solid docs and MIT license make it low-risk to try. Setup once, then watch your prompts level up.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.