peter/devclaw-gitea
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: refactor model selection to use developer tiers

Browse Source 
- Replaced raw model aliases with developer tiers (junior, medior, senior, qa) in dispatch and model selection logic.
- Updated `dispatchTask` to resolve models based on tiers and plugin configuration.
- Modified `selectModel` to return tier names instead of model aliases based on task description.
- Implemented migration logic for transitioning from old model aliases to new tier names in worker state.
- Added setup logic for agent creation and model configuration in `setup.ts`.
- Created shared templates for workspace files and instructions for DEV/QA workers.
- Enhanced project registration to scaffold role files based on developer tiers.
- Updated task management tools to reflect changes in model selection and tier assignment.
- Introduced a new `devclaw_setup` tool for agent-driven setup and configuration.
- Updated plugin configuration schema to support model mapping per developer tier.
This commit is contained in:
Lauren ten Hoor
2026-02-09 13:41:22 +08:00
parent 8a79755e4c
commit aa8e8dbd1b
16 changed files with 1162 additions and 257 deletions
Download Patch File Download Diff File Expand all files Collapse all files

102
docs/ONBOARDING.md
View File

@@ -9,9 +9,8 @@
| [`glab`](https://gitlab.com/gitlab-org/cli) or [`gh`](https://cli.github.com) CLI | Issue tracker provider (auto-detected from remote) | `glab --version` or `gh --version` |
| CLI authenticated | Plugin calls glab/gh for every label transition | `glab auth status` or `gh auth status` |
| A GitLab/GitHub repo with issues | The task backlog lives in the issue tracker | `glab issue list` or `gh issue list` from your repo |
| An OpenClaw agent with Telegram | The orchestrator agent that will manage projects | Agent defined in `openclaw.json` |
## Setup steps
## Setup
### 1. Install the plugin
@@ -26,35 +25,38 @@ openclaw plugins list
# Should show: DevClaw | devclaw | loaded
```
### 2. Configure your orchestrator agent
### 2. Run setup
In `openclaw.json`, your orchestrator agent needs access to the DevClaw tools:
```json
{
"agents": {
"list": [{
"id": "my-orchestrator",
"name": "Dev Orchestrator",
"model": "anthropic/claude-sonnet-4-5",
"tools": {
"allow": [
"task_pickup",
"task_complete",
"task_create",
"queue_status",
"session_health",
"project_register"
]
}
}]
}
}
```bash
openclaw devclaw setup
```
The agent needs the six DevClaw tools. Session management (`sessions_spawn`, `sessions_send`) is **not needed** — the plugin handles session creation and task dispatch internally via OpenClaw CLI. Workers (DEV/QA sub-agent sessions) also use `task_complete` and `task_create` directly for atomic self-reporting.
The setup wizard walks you through:
### 3. Register your project
1. **Agent** — Create a new orchestrator agent or configure an existing one
2. **Developer team** — Choose which LLM model powers each developer tier:
- **Junior** (fast, cheap tasks) — default: `anthropic/claude-haiku-4-5`
- **Medior** (standard tasks) — default: `anthropic/claude-sonnet-4-5`
- **Senior** (complex tasks) — default: `anthropic/claude-opus-4-5`
- **QA** (code review) — default: `anthropic/claude-sonnet-4-5`
3. **Workspace** — Writes AGENTS.md, HEARTBEAT.md, role templates, and initializes memory
Non-interactive mode:
```bash
# Create new agent with default models
openclaw devclaw setup --new-agent "My Dev Orchestrator" --non-interactive
# Configure existing agent with custom models
openclaw devclaw setup --agent my-orchestrator \
--junior "anthropic/claude-haiku-4-5" \
--senior "anthropic/claude-opus-4-5"
```
### 3. Add the agent to the Telegram group
Add your orchestrator bot to the Telegram group for the project. The agent will now receive messages from this group and can operate on the linked project.
### 4. Register your project
Tell the orchestrator agent to register a new project:
@@ -83,14 +85,14 @@ The agent calls `project_register`, which atomically:
"issueId": null,
"startTime": null,
"model": null,
"sessions": { "haiku": null, "sonnet": null, "opus": null }
"sessions": { "junior": null, "medior": null, "senior": null }
},
"qa": {
"active": false,
"issueId": null,
"startTime": null,
"model": null,
"sessions": { "grok": null }
"sessions": { "qa": null }
}
}
}
@@ -101,10 +103,6 @@ The agent calls `project_register`, which atomically:
**Finding the Telegram group ID:** The group ID is the numeric ID of your Telegram supergroup (a negative number like `-1234567890`). You can find it via the Telegram bot API or from message metadata in OpenClaw logs.
### 4. Add the agent to the Telegram group
Add your orchestrator bot to the Telegram group for the project. The agent will now receive messages from this group and can operate on the linked project.
### 5. Create your first issue
Issues can be created in multiple ways:
@@ -123,7 +121,7 @@ The agent should call `queue_status` and report the "To Do" issue. Then:
> "Pick up issue #1 for DEV"
The agent calls `task_pickup`, which selects a model, transitions the label to "Doing", creates or reuses a worker session, and dispatches the task — all in one call. The agent just posts the announcement.
The agent calls `task_pickup`, which assigns a developer tier, transitions the label to "Doing", creates or reuses a worker session, and dispatches the task — all in one call. The agent just posts the announcement.
## Adding more projects
@@ -131,18 +129,50 @@ Tell the agent to register a new project (step 3) and add the bot to the new Tel
Each project is fully isolated — separate queue, separate workers, separate state.
## Developer tiers
DevClaw assigns tasks to developer tiers instead of raw model names. This makes the system intuitive — you're assigning a "junior dev" to fix a typo, not configuring model parameters.
| Tier | Role | Default model | When to assign |
|------|------|---------------|----------------|
| **junior** | Junior developer | `anthropic/claude-haiku-4-5` | Typos, single-file fixes, CSS changes |
| **medior** | Mid-level developer | `anthropic/claude-sonnet-4-5` | Features, bug fixes, multi-file changes |
| **senior** | Senior developer | `anthropic/claude-opus-4-5` | Architecture, migrations, system-wide refactoring |
| **qa** | QA engineer | `anthropic/claude-sonnet-4-5` | Code review, test validation |
Change which model powers each tier in `openclaw.json`:
```json
{
"plugins": {
"entries": {
"devclaw": {
"config": {
"models": {
"junior": "anthropic/claude-haiku-4-5",
"medior": "anthropic/claude-sonnet-4-5",
"senior": "anthropic/claude-opus-4-5",
"qa": "anthropic/claude-sonnet-4-5"
}
}
}
}
}
}
```
## What the plugin handles vs. what you handle
| Responsibility | Who | Details |
|---|---|---|
| Plugin installation | You (once) | `cp -r devclaw ~/.openclaw/extensions/` |
| Agent + workspace setup | Plugin (`devclaw setup`) | Creates agent, configures models, writes workspace files |
| Label setup | Plugin (`project_register`) | 8 labels, created idempotently via `IssueProvider` |
| Role file scaffolding | Plugin (`project_register`) | Creates `roles/<project>/dev.md` and `qa.md` from defaults |
| Project registration | Plugin (`project_register`) | Entry in `projects.json` with empty worker state |
| Agent definition | You (once) | Agent in `openclaw.json` with tool permissions |
| Telegram group setup | You (once per project) | Add bot to group |
| Issue creation | Plugin (`task_create`) | Orchestrator or workers create issues from chat |
| Label transitions | Plugin | Atomic label transitions via issue tracker CLI |
| Model selection | Plugin | LLM-selected by orchestrator, keyword heuristic fallback |
| Developer assignment | Plugin | LLM-selected tier by orchestrator, keyword heuristic fallback |
| State management | Plugin | Atomic read/write to `projects.json` |
| Session management | Plugin | Creates, reuses, and dispatches to sessions via CLI. Agent never touches session tools. |
| Task completion | Plugin (`task_complete`) | Workers self-report. Auto-chains if enabled. |