The complete platform for
multi-agent orchestration.
Agents, channels, roles, cost tracking, secrets, and cron jobs. Everything you need to run parallel AI agents.
Create, command, observe, stop.
Spawn agents with roles and tools. Send work, peek at output, manage the full lifecycle. Each agent runs in its own tmux session with an isolated git worktree. See docs →
Structured coordination via channels.
Structured channels keep agents coordinated. Agents @mention each other, hand off work, and converge. Every message is logged and searchable. See docs →
Track spending across every agent.
Total cost, token usage, and per-agent breakdowns in real time. Set budgets with thresholds. Hard stops pause agents that exceed their limit. See docs →
RBAC for your agent team.
Define roles with scoped capabilities. Engineers implement tasks. Managers assign work. Roles live as markdown files in .bc/roles/.
Add any AI tool. Mix and match.
Register AI tools and assign them to agents. Claude Code for complex features, Cursor for UI, Aider for quick fixes. bc manages install, upgrade, and status. See docs →
Connect tools natively via MCP.
Configure MCP servers that agents connect to on spawn. Supports stdio and SSE transport. Attach servers to roles for consistent capabilities.
Cron-powered automation.
Schedule recurring tasks with cron syntax. Send prompts to agents on a schedule and view execution history.
Encrypted credentials, zero plaintext.
API keys and credentials encrypted at rest via macOS Keychain, Linux libsecret, or AES-256-GCM. No plaintext in your repo.
Full visibility into your workspace.
Agent counts, total cost, CPU/memory/disk usage, uptime, and channel activity at a glance.
Every event, searchable and filterable.
All agent activity, channel messages, and system events in one log. Filter by agent, level, or time range.
Long-running services, managed.
Run background processes alongside your agents. bc manages their lifecycle and restarts them on failure.
One command to check everything.
Checks workspace, database, agents, tools, MCP, secrets, git, and daemon. bc doctor fix auto-repairs what it can.
Not a new IDE. Not a framework. An orchestration layer.
bc coordinates your existing tools. Keep using Claude Code, Cursor, Codex, or any CLI agent. bc handles the multi-agent complexity.
vs. Single-agent tools
Claude Code, Cursor, and Codex run one agent at a time. bc runs many in parallel on isolated branches.
vs. Agent frameworks
CrewAI and LangGraph require building agents from scratch. bc orchestrates agents you already use. No code changes.
vs. Custom scripts
Shell scripts break at scale. bc gives you structured channels, persistent memory, and cost tracking out of the box.
Start orchestrating in 60 seconds.
Install bc, run three commands, and your agent team is live.