Autonomous Agents
Autonomous agents use brainclaw via CLI skills with --profile compact. Each agent reads from and writes to the shared coordination layer for persistent memory across runs.
Structured project memory for autonomous task-based execution.
OpenClaw is a task-based autonomous agent that uses brainclaw as a skill for persistent, structured project memory. Each run reads and writes to the shared coordination layer, keeping autonomous tasks aligned with the project's constraints, plans, and known traps without re-introducing context every time.
Autonomous skills-first agent with native MCP access.
Hermes Agent is an autonomous task-based agent from Nous Research with native MCP client support and a skills system. brainclaw configures Hermes as a Tier B autonomous surface today: machine-level MCP through ~/.hermes/config.yaml plus the shared .agents/skills/brainclaw/SKILL.md skill, so Hermes can read plans, claims, decisions, and traps from the same coordination layer as the rest of the team.
Messaging agent — WhatsApp, Telegram, Slack.
NanoClaw connects brainclaw's coordination layer to messaging channels like WhatsApp, Telegram, and Slack, keeping teams informed without leaving their inbox. Project decisions, plan updates, and trap captures can flow as native messages while brainclaw handles persistence.
NVIDIA enterprise agent stack bridge.
NemoClaw connects brainclaw to NVIDIA's enterprise agent stack so GPU-accelerated agents can participate in coordinated multi-agent flows. Agents using NVIDIA's reasoning runtimes inherit the same shared memory, constraints, and audit trail as the rest of the team.
Edge / IoT agent — Go, <10 MB RAM footprint.
PicoClaw is a scheduled Go-based agent under 10 MB RAM, designed for periodic coordination on edge devices and IoT infrastructure. It reads brainclaw's skill output and emits coordination events without requiring a full agent runtime.
Ultra-lightweight Rust agent — 20+ messaging channels.
ZeroClaw is an ultra-lightweight Rust agent supporting 20+ messaging and integration channels for high-throughput multi-agent flows. It reads brainclaw's skill output and bridges coordination state across channel surfaces without runtime overhead.
Autonomous agents use brainclaw via CLI skills with --profile compact for short sessions and constrained resources.