moo.agent.brain.chain

Token chain text processing — pure function over server output. See docs/source/explanation/agent-internals.md (Token Chain Mechanics) for the full design narrative. tests/test_brain_chain.py is the table-driven regression net.

Functions

process_server_text(text, state, config[, now])

Classify a server output line and produce the resulting ChainActions.

Classes

ChainActions([scripts, scripts_prepend, ...])

Side effects the Brain should apply after chain processing.