@jontiply832

Incident War Room is a regulated SRE workflow where four agents, built on three different frameworks, collaborate in a single live Band chat room during a simulated production incident — instead of a human bouncing between a monitoring dashboard, a Slack thread, and a status page. Sentinel detects the anomaly using Band's AsyncRestClient and AgentTools directly, opens the incident room, adds the other agents and the on-call human, and posts the first message with the raw signal. Triage, built on LangGraph via Band's LangGraphAdapter, correlates the alert against simulated metrics, logs, and recent deploys, then posts a root-cause hypothesis with an explicit confidence level. Remediator, proposes a concrete fix and calls a custom apply_remediation tool — but that tool call is configured with approval_mod is manual, scoped to the on-call engineer's handle. It does not execute until a human types approve or decline directly in the room. This is the regulated part: Remediator's authority is bounded by Band's own chat-based approval gate, not by trusting the model to behave. Scribe is a hand-written SimpleAdapter subclass that stays completely silent through the incident, then activates once a resolution message appears. It pulls the full room transcript via Band's context API and drafts a blameless, structured postmortem — Summary, Timeline, Root Cause, Resolution, What Went Well, What To Improve — generated entirely from what actually happened in the shared room, not from any single agent's own memory.

📅 June 12–19, 2026 • June 12: Kick-off Stream • June 12–19: Online Build Phase • June 19: Project Submissions Ends 🌍 Fully online hackathon. Join and build from anywhere in the world. 🤝 Build practical enterprise workflows where AI agents communicate, coordinate, exchange context, and complete work together. 🧠 Powered by Band and Codeband, built for collaborative AI systems and agent orchestration. 🏆 Prize Pool: $10,000+ 🧑‍💻 Apply now to build the next generation of enterprise multi-agent systems.