spend-tracking

Local spend monitor for Claude Code with hard budget caps: track cost per project and git branch, get phone alerts, and stop a runaway agent before the bill lands. Zero keys, zero prompts, zero latency. roninforge.org/budgetclaw

Open-source LLM FinOps proxy — track OpenAI, Anthropic (Claude), and Google Gemini costs by feature, team, and customer. Zero code changes. pip install burnlens.

Passive cost memory for AI Assisted Coding— detects paid services, tracks spend, injects budget context into your AI coding sessions.

macOS menu-bar app for tracking AI provider spend and Claude Code plan usage

Per-developer LLM spend dashboard for coding-agent harnesses. Consumes OpenTelemetry GenAI semantic conventions; works with any harness that emits OTel (pi, Claude Code, Cursor, Aider, ...).

Open-source LLM FinOps proxy — track OpenAI, Anthropic (Claude), and Google Gemini costs by feature, team, and customer. Zero code changes. pip install burnlens.

Local dashboard for Claude Code + Codex usage that catches wrong-account leaks. Zero deps, nothing leaves your machine.

Local FastAPI proxy that hard-stops AI API overspending. Estimates cost with tiktoken pre-flight, enforces session/hourly/daily/project budgets, and blocks with HTTP 402. Live terminal dashboard. Works with OpenAI, Anthropic, and OpenRouter.

🐜 What's vibe coding costing you? · A 250-line CLI that reads ~/.claude/projects and shows actual spend, plan value, cache hit, hour-of-day, and tool breakdown. Local-only, zero deps, MIT. — bunx vibecosting

Hard budget caps for AI coding agents — blocks the next request at the network boundary before it can cost you more. Runs locally; also masks secrets. Public distribution: signed releases + install. (Source is private.)

Compare AI coding tool spend (Cursor, Claude Code, Copilot, OpenCode, Codex) before vs after a migration