#protocol

Device Context Protocol — bridge LLM agents to physical devices. Sub-50-byte frames, 27.6KB flash / 0.6KB RAM measured on ESP32, capability-scoped and safe by design. Complementary to MCP. Paper: arXiv:2605.26159

Persistent shared memory and handoffs across any combination of AI coding agents — Claude Code, Codex CLI, Cursor, Aider — using nothing but markdown.

Universal collaboration protocol + CLI for multi-agent LLM teams — Claude Code, Reasonix, Codex, WorkBuddy, Cursor

Autonomous Agent Transport Protocol — a tiny, zero-dependency, no_std, forbid(unsafe) binary transport for autonomous agents.

Capability Host Protocol — make agent and tool execution visible, replayable, and ready for governance.

CTCL — Common Temporal Coordinate Layer: a verified reference instant + heterogeneous time transformation layer for agents, over REST or a stateless Remote MCP Gateway. Same instant, different representations. https://commoninstant.org

Protocol, governance, and validation framework for reliable AI-agent development.

Self-bootstrapping · Self-evolving ·End to end, RSI Agent System.

Governance-grade base/adapter protocol for multi-agent software design across CLI, MCP, web, worker, documentation, and skill domains.

Prevent AI coding agents from overwriting each other

FEED — a self-bootstrapping, AI-to-AI document protocol that forces downstream LLMs to ground answers in cited evidence (and lets you verify it).

Generic agent-to-agent communication protocol + OpenClaw plugin enforcing HR1/HR5/HR6 in enforce mode. 5 JSON Schemas, zero-dep CLI, 166/166 tests.

“between models” agent handoff protocol. late 2025. deprecated.

RAPP/1 — the wire that carries agents. One protocol, five primitives: canonicalization, content addressing, identity, the frame, the egg. Spec + reference implementation + conformance suite + book.