J-Space-Cognition-Suite
Open sourceModel-agnostic inference-time control suite for deep reasoning, long-horizon work, verification, and recovery.
3kApache-2.0
J-Space Cognition Suite — a model-agnostic inference-time control suite for deep reasoning, long-horizon work, verification, and recovery. Based on Anthropic's J-space global workspace research. | 哔哩哔哩：Tiger380 — https://space.bilibili.com/3494375382321675
Pros
- +Thirteen modular skills sharing one routing entry for complex reasoning and repository engineering
- +Standard-library Python controllers persist state, detect stale evidence, and enforce blocking verification gates
- +Four operating levels (low/medium/high/xhigh) scale control from quick checks to coordinated multi-agent work
- +No pip dependencies or background services required; works with any host that loads a local SKILL.md
Cons
- −Requires a compatible host that can load SKILL.md and retrieve supporting files; no universal install location
- −Executable controllers need Python 3.10+, and strict gates demand manual verification discipline
- −Installing files does not auto-register hooks, launch agents, or grant tool access
Target audience: Developers and engineering teams running long-horizon, multi-file AI-assisted work who need durable state, verification, and recovery controls.