day50-dev

day50-dev / capit

Public

Capping LLM spends per agent and per provider

agentic-ai agents key-management llm llm-inference
14
1
100% credibility
Found Mar 23, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Capit is a tool for creating spending-limited credentials for AI services and automatically configuring them into AI agent applications like Claude Code and Cursor, with local storage and backups.

How It Works

1
😰 Worry about AI costs

You hear about Capit when you're nervous that your AI helpers might spend too much money without limits.

2
📥 Get Capit ready

You easily add Capit to your computer so you can start setting budgets right away.

3
🔗 Link your AI service

You connect one of your AI services by sharing its main access details once, safely stored on your own machine.

4
💰 Set a budget for a helper

You pick one of your AI coding buddies, like Claude or Cursor, and give it a specific spending limit, like $5.

5
👀 Preview the changes

Capit shows you exactly what will update in your helper's settings with a safe preview, and you give the okay.

6
✅ Budget is set

Your AI helper now has the new limited budget, and your old settings are safely backed up just in case.

😴 Spend worry-free

Now you can use your AI helpers confidently, knowing they can't go over your set budget.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 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 capit?

Capit is a Python CLI for capping LLM spends per agent and per provider, like giving Claude Code or Cursor a $5 OpenRouter key that can't exceed budget. Run `capit openrouter 5.00 --agent claude` to generate an enforced key, auto-install it with config backups, and list/delete via `capit --keys list`. It handles soft capping LLM costs locally—no servers, just provider APIs for hard limits.

Why is it gaining traction?

Unlike manual key juggling or vague budgets, capit delivers one-command per-agent caps with diffs, backups, and usage tracking across providers like OpenRouter and AIHubMix. Devs love the "sleep soundly" hook: rogue agents hit walls at $5, not $500. Extensible for custom agents/providers keeps it fresh without bloat.

Who should use this?

AI power users tweaking Cursor/Windsurf for side projects, teams deploying OpenClaw agents on shared credits, or solo devs capitalizing LLM experiments without surprise bills. Perfect for per-provider spend control in multi-agent setups.

Verdict

Grab it if you're burning LLM cash—solid for niche agent workflows, MIT-licensed, but 14 stars and 1.0% credibility scream early alpha: thin docs, no tests visible. Test small before production.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.