Jarvis Voice

<p align="center"> <img src="docs/assets/logo.png" alt="TinkerClaw" width="500"> </p> <h1 align="center">TinkerClaw</h1> <p align="center"> <strong>An OpenClaw fork for people who check their token spend before breakfast.</strong> </p> <p align="center"> <a href="https://github.com/openclaw/openclaw"><img src="https://img.shields.io/badge/fork%20of-OpenClaw-5865F2?style=for-the-badge" alt="Fork of OpenClaw"></a> <a href="https://github.com/globalcaos/tinkerclaw/commits/main"><img src="https://img.shields.io/badge/hundreds%20of-fork%20commits-orange?style=for-the-badge" alt="Hundreds of fork commits"></a> <a href="#-published-skills"><img src="https://img.shields.io/badge/skills-20+-green?style=for-the-badge" alt="20+ skills"></a> <a href="#-memory-research"><img src="https://img.shields.io/badge/papers-7-blueviolet?style=for-the-badge" alt="7 papers"></a> <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-blue?style=for-the-badge" alt="MIT License"></a> </p> --- <p align="center"> <img src="docs/assets/screenshot-3.png" alt="Token timeline — every bar is a turn, every color is a cost" width="750"> <br> <em>Every bar is a turn. Every color is a cost. The spike at 14:02? A 40K-token tool result nobody asked for.</em> </p> > The Tinker UI's token visualization was inspired by [Mission Control](https://github.com/crshdn/mission-control) (context anatomy dashboard) and [ClawMetry](https://github.com/vivekchand/clawmetry) (real-time agent observability). Both are excellent standalone tools for OpenClaw — we folded their ideas into a single embedded panel. --- ## The Problem You ran Opus for 20 minutes. It felt productive. Then you checked the bill and discovered that "productive session" cost $23. The worst part? $15 of that was context bloat — workspace files you forgot were injected, tool results the model never referenced, conversation history from six topics ago still sitting in the window. You didn't overspend. You **overloaded**. And you had no way to see it happening. Most people find out three days later. The observant ones set a budget alert after it's already too late. The really unlucky ones hit their API rate limit at 2 AM on a Tuesday and wake up to a dead agent and no idea why. TinkerClaw exists because we got the $23 bill. Then we got angry. Then we built a dashboard. Then we couldn't stop — memory consolidation, self-improving crons, dozens of published skills, 7 research papers on how agent memory should actually work. Because once you start tinkering, you don't stop. --- ## 🤝 Come Tinker With Us This fork moves fast, but it would move faster with more hands. We value people who **open PRs**, not issues. Who read the code before asking questions. Who break things on purpose to understand how they work. If that's you, we want you in the inner circle — direct access to the roadmap, early testing of experimental features, and co-authorship on whatever we build next. **Start anywhere:** fix a typo, improve a skill, add a test, or propose something wild. The bar is curiosity, not credentials. → [Open a PR](https://github.com/globalcaos/tinkerclaw/pulls) or [start a discussion](https://github.com/globalcaos/tinkerclaw/discussions) --- ## Won't This Fork Fall Behind? No. A nightly cron syncs upstream automatically, detects conflicts, and restores fork patches after every merge. Hundreds of commits ahead of vanilla OpenClaw and zero behind. When upstream pushes a breaking change, we know within hours — not weeks. --- ## What You Get ### 🔍 Tinker UI — See Why Sessions Get Expensive The Tinker UI is a command center embedded directly in OpenClaw. No separate install, no external service. <p align="center"> <img src="docs/assets/screenshot-4.png" alt="Full Tinker UI — chat, sessions, tool calls, streaming" width="800"> <br> <em>Chat interface with session switching, tool call inspection, and real-time streaming.</em> </p> - **Context treemap** — drill into what fills your 200K context window, from categories down to individual messages and raw text - **Response treemap** — see exactly how much of each response is text, thinking, tool calls, or tool results - **Timeline** — stacked bars per turn, spot the one that blew the budget - **Overseer graph** — catch stalled sub-agents before they burn money - **Cost dashboard** — per-provider usage with Claude's 5-hour rate-limit countdown <table> <tr> <td width="50%"> <img src="docs/assets/screenshot-2.png" alt="Context treemap — drill into token composition" width="100%"> <br><em>Context treemap: every block is tokens you're paying for.</em> </td> <td width="50%"> <img src="docs/assets/screenshot-5.png" alt="Treemap drilldown — tool result detail" width="100%"> <br><em>Drill into a single category. These tool results cost $0.81 each.</em> </td> </tr> </table> After `pnpm build`, visit **`http://localhost:18789/tinker/`** · Dev: `cd tinker-ui && pnpm dev` --- ### 🧠 Memory That Improves Overnight Every night, the agent reviews its day — not as a diary, but as an evolution loop. - **ENGRAM consolidation** — raw daily logs → structured knowledge files → entity and project indices - **Retrieval feedback** — tracks which memory search results actually got used, prunes what doesn't help - **Structured compaction** — preserves decisions, tradeoffs, and open questions while discarding noise Measured: 23.5KB injected context → 12KB. **49% smaller, zero quality loss.** <p align="center"> <img src="docs/assets/screenshot-1.png" alt="Models and sessions panel" width="600"> <br> <em>Multi-model dashboard: Claude, Gemini, and local models with live status.</em> </p> --- ### 🔄 Self-Improving Agents Each cron job carries a META file with its own instructions. After running, the agent reflects on what worked, updates the META, and the next run is better. No human needed. Day 1: mediocre. Day 30: genuinely useful. ### 🧹 Fork Maintenance on Autopilot - Nightly upstream sync with conflict detection - Post-merge workspace cleanup (catches 20KB bloat) - Fork patches auto-restored after conflicts - Hundreds of commits ahead, zero maintenance burden --- ## 📦 Published Skills > All on [ClawHub](https://clawhub.ai/u/globalcaos). Install any with `clawhub install globalcaos/<skill-name>`. > Skills sometimes get delisted from the marketplace — this list is the permanent record. ### 🎤 Voice & Personality | Skill | What it does | |-------|-------------| | [`jarvis-voice`](https://clawhub.ai/globalcaos/jarvis-voice) | Turn your AI into JARVIS. Voice, wit, and personality — the complete package. | ### 💬 Messaging & Channels | Skill | What it does | |-------|-------------| | [`whatsapp-ultimate`](https://clawhub.ai/globalcaos/whatsapp-ultimate) | 3-rule security gate — agent speaks only when spoken to, in the right chat, by the right person. | ### 📹 Media & Content | Skill | What it does | |-------|-------------| | [`youtube-ultimate`](https://clawhub.ai/globalcaos/youtube-ultimate) | Free transcripts, 4K downloads, video exploration — zero API quotas burned. | ### 💰 Cost & Token Management | Skill | What it does | |-------|-------------| | [`tinker-command-center`](https://clawhub.ai/globalcaos/tinker-command-center) | The dashboard above. Every token, every dollar, every context byte — real time. | | [`token-panel-ultimate`](https://clawhub.ai/globalcaos/token-panel-ultimate) | Multi-provider token tracking, budget alerts, REST API. | | [`token-efficiency-guide`](https://clawhub.ai/globalcaos/token-efficiency-guide) | Go from weekly limit on Tuesday to weekly limit on Sunday. 10 steps, one afternoon. | ### 🏢 Enterprise Integrations (Browser Relay) No API keys. No admin consent. Your authenticated browser session IS the API. | Skill | What it does | |-------|-------------| | [`outlook-hack`](https://clawhub.ai/globalcaos/outlook-hack) | Reads Outlook all day, drafts replies — won't send without approval. Code-enforced. | | [`teams-