NullRun is the enforcement layer for AI agents. Drop in a one-line decorator, set a budget, and your agent stops when cost spirals, loops, or risky tool calls — before the bill lands, not after.

• Stop runaway costs. Hard budget caps on every workflow. The agent halts the moment the cap is hit, not in next week's report.
• Catch loops and retry storms. Loops, retry storms, and over-eager tools are detected and blocked mid-run, not after the API bill does the talking.
• Kill from the dashboard. When something goes wrong, hit Kill and the agent stops within milliseconds — even mid-execution. WebSocket control plane, not polling.
• Run sensitive tools safely. Database writes, shell, file deletes — gate each one with an explicit policy. Approvals, allow-lists, full audit trail.

• 🟢 nullrun.io — managed control plane and dashboard in one place. Free Lite plan, no credit card.
• 🐍 pip install nullrun — Python SDK with one-line instrumentation for OpenAI, Anthropic, OpenAI Agents, LangGraph, and the rest.
• 📖 docs.nullrun.io — install, quickstart, concepts, and recipes.

• nullrun-sdk-python — the Python SDK (pip install nullrun).
• nullrun-docs — the documentation site.
• nullrun-examples — copy-paste-runnable examples for the most common agent frameworks.

• 🐛 Bug or feature? Open an issue in the relevant repo.
• 🔒 Security? See SECURITY.md — please don't file public issues.
• 💬 Anything else: support@nullrun.io.