# J-Space Cognition Suite

J-Space Cognition Suite V3.7 - AI cognitive-enhancement Skills based on Anthropic's J-space global workspace research. | 哔哩哔哩：Tiger380 (UID 3494375382321675) — https://space.bilibili.com/3494375382321675

Repository: https://github.com/Tiger3807861189/J-Space-Cognition-Suite-V3.7
Canonical: https://ross.abutalabs.com/products/j-space-cognition-suite
Language: Python
License: Apache-2.0
License Family: permissive
Topics: dsh-plugin, agent-skills, ai, ai-agent, ai-agents, claude-code, codex, cognitive-enhancement, deepseek, deepseek-harness, developer-tools, dsh, global-workspace, hermes-agent, j-space, opencode, react, tailwindcss, inference-time-control
Last push: 2026-08-23T07:27:40+00:00

## Health v2 (maintenance only)
Score: 69/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 67, longevity 3
- inputs: {"age_days": 43, "days_push": 10, "days_rel": 10, "gap_med": null, "n_releases_24m": 1}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3018, forks 216 (observed 2026-08-28T04:07:38.385464+00:00)

## What it is
A model-agnostic inference-time control suite packaged as an installable AI agent Skill (SKILL.md with modules, references, and Python verification scripts). It organizes an agent's working representations into a managed workspace for deep reasoning, long-horizon tasks, tool use, verification, and recovery without changing model weights.

## Use cases
- improve agent reasoning on long-horizon tasks
- add structured verification and recovery to AI agent workflows
- install a cognitive skill into Claude Code or Codex
- control LLM behavior at inference time without fine-tuning
- manage durable task state across multi-step agent work
- audit a repository while preserving its architecture

## When to choose
- you use a Skill-compatible AI host like Claude Code, Codex, or OpenCode and want structured reasoning control
- you need inference-time behavior shaping without retraining or fine-tuning a model
- you want a model-agnostic, selectively loaded skill with a single entry point

## When to avoid
- your AI host has no native Skill loader and you cannot use system/developer-instruction integration
- you expect weight-level or training-time model changes
- you need a standalone application or GUI rather than an agent-invoked skill

## Facets
- artifact type: plugin
- maturity: active
- function: agent-framework, prompt-engineering, llm-inference, developer-tools
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: python, cross-platform, cli
- tags: agent-skills, claude-code, codex, opencode, inference-time-control, global-workspace, cognitive-enhancement, skill-plugin, deepseek, model-agnostic, ai-agents

## Member repositories
- Tiger3807861189/J-Space-Cognition-Suite-V3.7 (main) score 69
- Tiger3807861189/GLM-5.3-Flash-J-Space-Capability-Realization-Report (docs) score 69

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:38.385464+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:30:02.245409+00:00, confidence not recorded.
  - readme: https://github.com/Tiger3807861189/J-Space-Cognition-Suite-V3.7 (fetched 2026-08-28T04:07:38.385464+00:00, sha 0628e83b67f0)
- Data as of 2026-08-30T08:39:29.467469+00:00.