Clavue 10.9.5 · China-built AI coding CLI

Dragon-grade execution for serious code.

Clavue is China-built software for serious engineering teams: direct repo tools, shell execution, provider routing, MCP resources, task tracking, worktrees, skills, and long-session recovery in one disciplined coding runtime.

Get started Read docs

China-built software龙Execution power with disciplined engineering

$ npx -y clavue@10.9.5
> Inspect the failing provider-routing test, make the smallest source fix, run the targeted test, then verify before reporting back.

Clavue
  ✓ read tests/provider-config-manager.test.mjs
  ✓ edited src/utils/providerProfiles.ts
  ✓ ran node --test tests/provider-config-manager.test.mjs
  ✓ verified npm run validate:repo

npxInstant first run

/providerRoute control center

/maoReview and rescue

Install and first run

Start from the terminal. Stay in the execution loop.

Run once with npx, install globally for daily use, or pin a release with the installer script.

npx

npx -y clavue@10.9.5 --version
npx -y clavue@10.9.5

npm global

npm install -g clavue@10.9.5
clavue install --force
clavue

curl

curl -fsSL https://unpkg.com/clavue@10.9.5/install.sh | bash -s -- 10.9.5

Chinese software · dragon spirit

A coding CLI with Chinese momentum and engineering discipline.

The dragon visual system represents focus, strength, continuity, and verified delivery. Provider compatibility is a runtime capability. Clavue remains the product and workflow identity.

Direct repo tools

Inspect files, search code, edit source, and run checks from the CLI instead of staying in a detached chat surface.

Provider routing control

Save profiles, assign model slots, validate reachability, and repair drift without making any provider the product identity.

Verification discipline

Clavue is designed to report concrete checks: targeted tests first, then the smallest broader gate that proves the change.

Native workflows

/provider, /mao, /team, /retro, task tracking, worktrees, skills, and MCP resources stay in one runtime loop.

Long-session recovery

Task state, compaction, and recovery behavior help extended coding sessions continue without restarting context from zero.

Team-ready operations

Inspect local team readiness, active configuration, permission posture, and rollout state before scaling usage.

Provider routing

Route models deliberately. Keep the workflow Clavue-native.

Configure official login, custom API profiles, compatible gateways, and model slots. Validate reachability and repair drift without changing your daily coding loop.

/provider
/provider list
/provider current
/provider doctor
/provider validate
/provider repair
/provider use gpt54-main

Active provider profile

Profile: gpt54-main
Primary: coding model
Small/Fast: helper lane
Plan/Explore: high-capability lane
Status: Validated

Set CLAVUE_API_DIALECT=openai_responses only when a gateway explicitly requires the Responses adapter.

Best use cases

Prompt Clavue for one reviewable unit, then verify it.

Fix a failing test

Inspect the failing test around provider routing, make the smallest source fix, run the targeted test, then run the relevant verification command before reporting back.

Read the failing test and nearby source
Edit only the necessary files
Run the targeted test first
Report changed files and verification results

Review a risky branch

/mao adversarial-review --base main provider routing and permissions

Compare against the base branch
Prioritize routing and permission regressions
Return file references and concrete risks

Configure a provider

Open provider setup, create a saved profile for this API route, validate it, and make it the default if validation passes.

Use clavue provider or /provider
Save the profile with a clear name
Run /provider validate
Use /provider doctor if behavior drifts

Team rollout

/team check

Inspect readiness and active config
Confirm provider profile conventions
Document verification expectations