Skip to content

CLM v0.1.0 First Public Release

Choose a tag to compare

View all tags
@ragul-rofi ragul-rofi released this 31 Mar 16:54
· 7 commits to main since this release
v0.1.0
5191338

CLM v0.1.0 — First Public Release

Real-time metacognitive middleware for LLM agents. One line to add to any agent loop.

Install

pip install clm-agent

Quickstart

from clm import CLM
clm = CLM(verbose=True)
result = clm.observe_raw(llm_output)

What's in this release

  • 4 cognitive load signals (branching, repetition, uncertainty, goal drift)
  • 3-zone intervention system with Amber escalation and abort safety
  • LangChain, OpenAI Agents, and generic loop adapters
  • no_embed mode for offline environments
  • Domain configs for medical, legal, and voice agents

Known limitations in v0.1

  • Default weights are heuristic — tune them for your domain
  • No async support yet (v0.2)
  • No CrewAI/AutoGen adapters yet (v0.2)

Full docs in README. Issues and usage reports welcome.

Assets 2
Loading