A model-agnostic inference-time agent control system packaged as a cross-platform Skill (one entry, nine modules, four references, optional standard-library controller). Inference-time only, no weight change. Repo is tagged deepseek-harness / dsh / dsh-plugin, but the README describes no native DeepSeek Harness runtime integration.
DSH integration
Ecosystem-related
Author-claimed
Safety audit
Unaudited
Last verified
2026-09-03
License
NOASSERTION
01What can it help you accomplish?
Run a model-agnostic inference-time control system for deep reasoning, long-horizon work, tool use, verification, and recovery inside an AI host
A packaged J-Space Skill with one entry point, nine selectively loaded modules, four supporting references, and an optional standard-library controller for durable task state
AI-host and coding-agent users who want cross-platform, low-friction agent control without changing model weights or training
02How to install into DeepSeek Harness
Prerequisites
- A user-level Skills directory provided by your AI host
- A Python 3 interpreter available on the host (commonly `python`, `python3`, or `py -3`) for the integrity check
Installation steps
- 01
Download or clone this repository.
- 02
Locate the user-level Skills directory used by your AI host.
- 03
Copy the complete `j-space/` directory into it so that the installed entry is `<skills-directory>/j-space/SKILL.md`.
- 04
Run the integrity check: `<python-command> <skills-directory>/j-space/scripts/verify_suite.py`.
- 05
Reload the host if it discovers Skills at startup.
Verify the integration
- Run `<python-command> <skills-directory>/j-space/scripts/verify_suite.py` and confirm the suite passes its integrity check.
03DSH integration and capability boundaries
The README documents J-Space as a model-agnostic inference-time Skill for cross-platform agent control; the repo is tagged deepseek-harness / dsh / dsh-plugin, but the README does not describe a native DeepSeek Harness (dsh) runtime integration.
Model-agnostic inference-time control
an agent's accessible working representations→controlled deep reasoning, long-horizon work, tool use, verification, and recovery
Packaged cross-platform Skill
the user-level Skills directory of an AI host→an installed `j-space/` Skill whose `SKILL.md` entry routes to nine modules, four references, and scripts
`SKILL.md` routes to relative paths under `modules/`, `references/`, and `scripts/`, so the `j-space/` directory must stay intactInference-time operation (no training change)
any model at inference time→behavior controlled at inference time while model weights and training remain unchanged
04Who is it for? When not to use it?
Good for
- AI-host and coding-agent users who want cross-platform, low-friction agent control without changing model weights or training
Not for
- The `j-space/` directory must remain intact: `SKILL.md` routes to relative paths under `modules/`, `references/`, and `scripts/`, so do not split or move subfolders when copying the Skill.
05Compatibility, maintenance and safety notes
- The `j-space/` directory must remain intact: `SKILL.md` routes to relative paths under `modules/`, `references/`, and `scripts/`, so do not split or move subfolders when copying the Skill.
- J-Space operates at inference time and model weights and training remain unchanged, so it cannot alter a model's underlying capabilities.
License NOASSERTION; repo topics include deepseek-harness, dsh, dsh-plugin, but README asserts no security audit
06Frequently asked questions
Does this plugin integrate natively with DeepSeek Harness (dsh)?
The README documents J-Space as a model-agnostic inference-time Skill for cross-platform agent control and does not describe a native DeepSeek Harness runtime integration. The repository is tagged deepseek-harness, dsh, and dsh-plugin, but no dsh CLI or dsh plugin runtime is described in the README, so it is treated as ecosystem-level.
How do I install J-Space?
Download or clone the repository, copy the complete `j-space/` directory into your AI host's user-level Skills directory so the entry is `<skills-directory>/j-space/SKILL.md`, then run the integrity check with a Python 3 interpreter and reload the host.
What does J-Space require at runtime?
A user-level Skills directory from your AI host and a Python 3 interpreter (commonly `python`, `python3`, or `py -3`) for the integrity check. It operates at inference time and does not change model weights or training.
Does J-Space modify my model?
No. J-Space operates at inference time; model weights and training remain unchanged. It is a control system for reasoning, tool use, verification, and recovery, not a fine-tuning or training step.
What do the deepseek-harness / dsh / dsh-plugin topics on the repo indicate?
These topics indicate the repo is catalogued as DSH-related, but the README only describes a host-agnostic Skill install and documents no DSH-specific runtime path. Until the README is updated, the integration level is ecosystem.
07Related DSH workflows
archify
by tt-a1i
Turn any idea, plan, or codebase into a beautiful interactive diagram. An agent skill for Claude Code, Codex, and more.
openviking
by volcengine
Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
nocobase
by nocobase
NocoBase is an open-source AI + no-code platform for building business systems fast. Instead of generating everything from scratch, AI works on top of production-proven infrastructure and a WYSIWYG no-code interface, so you get both speed and reliability.
learn-harness-engineering
by walkinglabs
Harness engineering beginner tutorial, from 0 to 1
08Data and sources
It is packaged as a Skill for cross-platform use, selective loading, and low-friction integration.
J-Space Cognition Suite is a model-agnostic inference-time control system for deep reasoning, long-horizon work, tool us…
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-09-03. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.