Skip to content

sportsclawBuild AI that understands live sports.

An open-source engine with keyless live data, market odds, and real-time game events built in. Ship chat bots, broadcast widgets, and odds trackers in minutes — not months of data plumbing.

sh
curl -fsSL https://sportsclaw.gg/install.sh | bash

macOS / Linux — verify the published SHA-256 before running. Windows: npm install -g sportsclaw-engine-core

┌───────────┐
│ LAL · 102 │
│ GSW ·  98 │
└───────────┘

Answers grounded in real data

Ask in plain language and get answers backed by live scores, standings, stats, odds, and news across 14 sports. The agent looks it up — it doesn't guess.

╭─────────╮
│ score?  │
╰─────────╯
  ╭─────────╮
  │ LAL 102 │
  ╰─────────╯

Discord & Telegram bots

Run a sports bot with rich embeds, buttons, polls, and image replies in one command. Your community asks; the bot answers with the real numbers.

   ▟█▙
  ▟███▙
  ▀▀▀▀▀
    

It tells you when things happen

Followers say "alert me about the Lakers" and get a message the moment the game starts, the lead changes, or it ends — no polling, no setup.

┌─────────┐
│      ▂▅│
│   ▂▅    │
│▂▅       │
└─────────┘

Live odds & prediction markets

Pull real-time odds from ESPN, Kalshi, and Polymarket, and run the betting math. Read-only by default — built to track, not trade.

┌────────┐
│   ▁▄   │
│ ▁▄██▄▁ │
└────────┘

Generate graphics on the fly

Ask for a matchday graphic and get one back, delivered straight into the chat. Vision works too — send a screenshot and ask about it.

┌────────┐
│ > help │
│ > _    │
└────────┘

Open source, runs anywhere

MIT-licensed TypeScript. Run it from the CLI, host it as a bot, or deploy in Docker. Bring your own model — Anthropic, OpenAI, Google, or any OpenAI-compatible endpoint (Azure AI Foundry, self-hosted NIM/vLLM).

┌─────────┐
│ turn 3  │
│ resumes │
└─────────┘

Delegate durable, resumable work

Connect a Machina pod and hand off long, multi-step tasks to its durable loop — it persists every turn and resumes after interruptions, async waits, or restarts. sportsclaw dispatches the work and reads the result back.

Hand it to your coding agent

Paste this into Claude Code, Cursor, or any coding agent — it reads the machine-readable doc index and builds from there:

text
Build a sports AI app with sportsclaw. First read https://sportsclaw.gg/llms.txt
for the full doc map, then follow https://sportsclaw.gg/getting-started/quickstart
to install and scaffold. sportsclaw gives you keyless live scores, standings, odds
and markets via sports-skills, plus one-command Discord and Telegram bots. Use the
docs at https://sportsclaw.gg for the CLI, data coverage, and deployment.

llms.txt lists every doc page as a fetchable URL, so the agent can crawl the whole reference on its own.

Day-one capabilities

Your agent ships with three built-in personas. Each learns from your chat — the more your users interact, the sharper it gets.

The Analyst

Prediction Markets

Drops odds and value bets into the chat. It learns which sports and markets your audience actually cares about and tailors its alerts.

The Scoreboard

Multi-Sport

Pulls live play-by-play and stats for any game. Over time it anticipates the matchups your audience follows and tracks them proactively.

The News Desk

Scheduled Reports

Curates morning injury reports and headlines, refining its daily feed based on the teams your chat interacts with most.

Open source under the MIT License.