hive-claude

Claude Code as a hive backend — both the terminal you watch and the headless loop you don't. It exposes Claude Code through the hive addon ports, so the same host code can drive an interactive Emacs terminal session or a fully headless Agent SDK run.

Coordinates

;; deps.edn
io.github.hive-agi/hive-claude {:mvn/version "0.1.11"}

Two backends, one addon

hive-claude.init is the IAddon implementation — a reify over a nil-railway pipeline, with zero compile-time hive-mcp dependencies; everything host-side resolves through requiring-resolve.

Terminal. hive-claude.terminal implements ITerminalAddon, bridging JVM-side ling management to the Emacs-side hive-claude elisp package by evaluating elisp through the Emacs client. The elisp is not hand-written: it is compiled from ClojureElisp sources under src/cljel/.

Headless. hive-claude.sdk.agentic-loop implements IAgenticLoop over the Claude Agent SDK:

• async start/abort lifecycle via sdk.lifecycle spawn/kill
• collect-response! with a CompletableFuture and a timeout
• cost tracking accumulated from SAA completion events
• transcript access from session observations
• mid-session constraints through budget guardrails

Layout

Path	Contents
src/cljel/hive_claude/	ClojureElisp sources — bridge, config, state, sync
src/clj/hive_claude/	Terminal adapter, elisp emission, addon init
src/clj/hive_claude/sdk/	Agent SDK lifecycle, execution, event loop, options, availability, SAA, phase compression

The Python-side Agent SDK dependencies (libpython-clj) are loaded lazily through requiring-resolve, so the terminal backend works on a JVM that has no Python environment at all.

License

AGPL-3.0-or-later.