DeepSeek Harness Plugin Hub

Publish and manage complete Harness Profiles. Discover Plugins for your next setup.

Explore

PluginsPresetsDocsNews

Community

Publish a pluginContactReport an issue

Resources

Plugin Hub on GitHubDeepSeek HarnessSystem statusPrivacy notice
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

Independent and unofficial. Not affiliated with, authorized by, or endorsed by DeepSeek.

Praxis — DSH Plugin for DeepSeek Harness
← Plugins
P

praxis

Praxis

DSH plugin praxis

The plugin will be installed here. Keep web if you are unsure.

npx -y @deepseek-ai/dsh plugin --profile web add github:Civitasv/praxis#efe5f0de2253d7a14885ec82434a25d9f3183dd7
READMECompatibilityVersions

Description

DSH plugin praxis

Compatibility and provenance

Praxis is published as praxis and currently resolves to version 0.0.0. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
any
Release source
github
Registry updated
10/6/2026

Versions

0.0.0stable
10/6/2026
Latest
0.0.0
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
Unavailable
Files
Unavailable
Surface
any
License
AGPL-3.0-only
Source
github
GitHub
★ 1
Weekly downloads
0
Last push
10/7/2026
View source ↗
README badge

Click the badge to copy Markdown for your README.

Do you maintain this Plugin?Claim benefit · Priority security scan

Verify the GitHub repository declared in package.json to manage this listing. After you claim it, Hub will prioritize a security scan of the current version and publish the result when it passes.

Claim this Plugin →
Report an issue
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in

README

Praxis

Praxis

Build with AI. Develop taste.

A plugin for developing engineering taste while coding with AI.

Codex · Cursor · CodeBuddy · DeepSeek Harness

English · 简体中文


Philosophy

AI makes implementation cheap. Taste is still earned.

Fast output is not the same as good engineering. If AI makes every meaningful choice before you understand it, you may ship faster today without getting better at choosing tomorrow.

Praxis extends the coding agent you already use. Routine work stays fast. When a decision can shape the system, Praxis slows down just enough to expose the reasoning, alternatives, tradeoffs, and consequences.

That attention compounds. The more good decisions you understand, the faster you can recognize them yourself.

Slower at the right moment is faster over time.

What using Praxis feels like

Without Praxis:

You     Add persistent chat history.

Agent   I’ll use the database as the source of truth
        and Redis as a cache.

        ✓ Implemented

With Praxis:

You     Add persistent chat history.

Agent   I’ll use the database as the source of truth
        and Redis as a cache.

        This choice matters:
        history is durable product state, so a Redis restart
        should not erase it.

        The tradeoff is an extra durable write path,
        but Redis stays disposable.

You     I don't know what "source of truth" means.

Agent   It means the database owns the canonical history.
        Redis can make access faster, but losing Redis
        should not lose messages.

You     Got it.

Agent   ✓ Implemented
        ✓ Verified: history survived a Redis restart.

Praxis keeps the learning inside the same workflow, so the decision becomes part of your own taste instead of disappearing into the implementation.

What Praxis helps you build

Engineering taste is the ability to recognize better choices earlier.

It grows by connecting decisions to consequences:

compare
→ choose
→ implement
→ observe
→ internalize
→ choose better next time

Praxis does not try to slow down the whole task. It spends attention where understanding compounds, so future decisions become faster, better, and more independent.

Install

Requires Python 3.10+.

Install Praxis:

python3 -m pip install --user 'git+https://github.com/Civitasv/praxis.git'

On Windows, replace python3 with py.

Then add Praxis to the coding agent you use.

Codex
codex plugin marketplace add Civitasv/praxis
codex plugin add praxis@praxis

Enable it in the project:

$praxis:praxis-enable
Cursor
mkdir -p ~/.cursor/plugins/local
git clone --depth 1 https://github.com/Civitasv/praxis.git ~/.cursor/plugins/local/praxis

Restart Cursor, then run:

/praxis enable
DeepSeek Harness / Cordis

Replace web with your profile name.

dsh plugin --profile web add github:Civitasv/praxis
dsh --profile web

Then run:

/praxis enable
CodeBuddy
codebuddy plugin marketplace add Civitasv/praxis --name praxis && codebuddy plugin install praxis@praxis

Then run:

/praxis:enable

Once enabled, work normally. Praxis only slows down decisions worth learning from.

Commands

AgentEnableDisableStatus
Cursor/praxis enable/praxis disable/praxis status
DeepSeek Harness/praxis enable/praxis disable/praxis status
CodeBuddy/praxis:enable/praxis:disable/praxis:status
Codex$praxis:praxis-enable$praxis:praxis-disable$praxis:praxis-status

Update Praxis

Update Praxis:

python3 -m pip install --user --upgrade 'git+https://github.com/Civitasv/praxis.git'

Then refresh the integration you use.

Codex
codex plugin marketplace upgrade praxis
codex plugin add praxis@praxis

Open a new chat.

Cursor
git -C ~/.cursor/plugins/local/praxis pull --ff-only

Restart Cursor.

DeepSeek Harness / Cordis
dsh plugin --profile web update praxis
dsh --profile web
CodeBuddy
codebuddy plugin marketplace update praxis && codebuddy plugin update praxis@praxis

Then reload plugins:

/reload-plugins

Status

Praxis is currently alpha.

Current integrations: Codex, Cursor, CodeBuddy, and DeepSeek Harness / Cordis. Distribution is currently source-based.

License

Praxis is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0-only).

You may use, modify, and redistribute Praxis under the terms of the AGPL. Its copyleft requirements include source-availability obligations for covered modifications, including when a modified version is offered for use over a network. See LICENSE for the full terms.

Documentation

Implementation details live under Docs/:

  • Architecture overview
  • Tutor model
  • Harness integrations
  • Feature specifications
  • Development and validation

Contributors should start with Code.md and State.md.