Find the ghost tokens. Fix them. Survive compaction. Avoid context quality decay.

Local Emperor-style AI agent with Vue WebUI, multi-provider LLMs, streaming chat, tools, skills, memory, and token telemetry.

A multi-user Claude Code CLI based Coworker with Multi Agent Orchestrator, Multi model support including Ollama models, Advanced Functional skills, Channel integration for Chat and text messaging platform running completely in docker containers with an Admin module for complete visibility and observability

High-precision quota and token monitoring for AI Agents. Real-time analytics and usage tracking for Claude, Codex, Gemini, and Copilot to prevent API exhaustion and optimize LLM budgets.

Claude Code usage dashboard in your macOS menu bar — costs, tokens, plan limits, trends, multi-machine sync

Local monitoring center for Codex CLI sessions: rolling 5-hour quota, token usage, estimated cost, and a NOC-style dashboard.

A privacy-first, fully local Hebrew/RTL dashboard for analyzing Claude Code token usage, costs, prompts, tools, skills, and sessions. Reads JSONL transcripts from ~/.claude/projects/, stdlib-only Python, no telemetry, no outbound HTTP.

Track, visualize, and optimize LLM API spending. Monitor OpenAI & Anthropic costs per feature, detect waste, suggest savings. Zero-config Python profiler.

Governance-first Codex agent skillset/skillpack for better results, including: workflow automation, requirement adherence, regression prevention, testing strategy, policy enforcement, documentation sync, interface design, and UX-focused execution.

Real-time budget monitor dashboard for Claude Code token usage

Local viewer for Claude Code logs with CLI dashboard and token usage tracking

One line to instrument your agent and capture every event in an immutable, queryable audit trail.

Atelier is an open-source reasoning runtime for coding agents, helping them reuse procedures across coding agents, make tool calling efficient, recover from failures, and avoid repeating mistakes. This is the only runtime you would need to make all you coding agents cheap.

G4 Claw Counter — Windows desktop widget tracking AI token usage & costs across Claude Code, Claude Desktop, OpenAI Codex, and more. Zero AI tokens consumed. Shows today's cost, multi-provider support, user-editable pricing.

Track every dollar your AI spends. Transparent proxy + real-time dashboard for any LLM API. One line change, works with any language.

Local Python analyzer for Claude session transcripts with token usage reports and prompt insights.

Star Tokens: local token/cost dashboard for Codex, Claude Code, and Antigravity.

the audit kit for finding hidden token leaks in openclaw stacks. ships with the consultant playbook to sell audits as a service.

Claude Code plugin: /tokens:tokens — historical token usage & cost breakdown by model, project, date. Install: claude plugin marketplace add PeterCang/claude-tokens-plugin

📊 Multi-account analytics for Amazon Bedrock. Hub + Spoke architecture aggregates invocation logs across AWS accounts into DynamoDB; NiceGUI WebUI shows token usage, cost breakdown, latency, and TPOT in real time.