mirror of
https://github.com/openclaw/openclaw.git
synced 2026-08-17 08:02:12 -06:00
Peter Steinberger
116 lines
7.9 KiB
Markdown
116 lines
7.9 KiB
Markdown
---
|
|
summary: "What experimental flags mean in OpenClaw and which ones are currently documented"
|
|
title: "Experimental features"
|
|
read_when:
|
|
- You see an `.experimental` config key and want to know whether it is stable
|
|
- You want to try preview runtime features without confusing them with normal defaults
|
|
- You want one place to find the currently documented experimental flags
|
|
---
|
|
|
|
Experimental features are preview surfaces behind explicit flags. They need more real-world mileage before they get a stable default or a long-lived contract.
|
|
|
|
- Off by default unless a doc describes a narrow automatic setup rule.
|
|
- Shape and behavior can change faster than stable config.
|
|
- Prefer a stable path when one already exists.
|
|
- Roll out broadly only after testing in a smaller environment first.
|
|
|
|
## Currently documented flags
|
|
|
|
| Surface | Key | Use it when | More |
|
|
| ------------------- | --------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------- |
|
|
| Local model runtime | `agents.defaults.experimental.localModelLean`, `agents.entries.*.experimental.localModelLean` | A smaller or stricter local backend chokes on OpenClaw's full default tool surface | [Local Models](/gateway/local-models) |
|
|
| Codex harness | `plugins.entries.codex.config.appServer.experimental.sandboxExecServer` | You want native Codex app-server 0.143.0 or newer to target an OpenClaw sandbox-backed exec-server instead of disabling Code Mode | [Codex harness reference](/plugins/codex-harness-reference#sandboxed-native-execution) |
|
|
| Code Mode | `tools.codeMode.enabled` | You want compact code-orchestrated access to a hidden OpenClaw tool catalog | [Code Mode](/tools/code-mode) |
|
|
| Cloud workers | `cloudWorkers.desktop` | You want to watch or control desktop-capable cloud worker environments from the Control UI | [Cloud Worker Desktop](/gateway/cloud-workers#desktop-interactive) |
|
|
| Swarm | `tools.swarm.enabled` | You want Code Mode scripts to orchestrate bounded groups of sub-agents in parallel | [Swarm](/tools/swarm) |
|
|
|
|
## Control UI Labs
|
|
|
|
Open **Settings → Agents & Tools → Labs** to manage experiments that have a
|
|
Control UI switch. Enabling or disabling a lab patches the canonical Gateway
|
|
config immediately; the page shows a restart hint only when a feature requires
|
|
one.
|
|
|
|
The currently shipped Labs entries are Code Mode, Swarm, Tool Search,
|
|
Tool-loop detection, Lean tools for local models, Message audit metadata, and
|
|
Cloud Worker Desktop. Message audit metadata and Cloud Worker Desktop require a
|
|
Gateway restart; the other switches normally take effect for future agent runs
|
|
without restarting.
|
|
|
|
## Local model lean mode
|
|
|
|
`agents.defaults.experimental.localModelLean: true` drops heavyweight optional tools from the agent's direct surface every turn: `browser`, `cron`, `message`, `image_generate`, `music_generate`, `video_generate`, `tts`, and `pdf`. Explicitly allowed or delivery-required tools remain available, though Tool Search may catalog them instead of exposing them directly. Lean mode also defaults plugin/MCP/client catalogs to structured Tool Search (`tool_search`, `tool_describe`, `tool_call`) when `tools.toolSearch` is not already set. Use `agents.entries.*.experimental.localModelLean` to scope this to one agent.
|
|
|
|
During onboarding, a verified `ollama` or `lmstudio` inference route automatically sets `agents.defaults.experimental.localModelLean: true` when that value is absent. OpenClaw records that the setting came from onboarding, so a later verified non-local route lifts only the automatic setting. An explicitly configured `true` or `false` is preserved. Other self-hosted and OpenAI-compatible providers are not inferred from model names or URLs.
|
|
|
|
If you already tune Tool Search globally, OpenClaw leaves that config alone. Set `tools.toolSearch: false` to opt out of the lean-mode Tool Search default.
|
|
|
|
In structured `tools` mode, lean runs keep `exec` directly visible beside the Tool Search controls so coding-tuned local models can still choose their familiar shell path. This changes schema visibility only: normal tool policy, sandboxing, and exec approvals still apply. Explicit `code` and `directory` modes keep their normal compaction behavior.
|
|
|
|
### Why these tools
|
|
|
|
These tools have the largest descriptions, broadest parameter shapes, or highest chance of distracting a small model from the normal coding and conversation path. On a small-context or stricter OpenAI-compatible backend that is the difference between:
|
|
|
|
- Tool schemas fitting the prompt vs. crowding out conversation history.
|
|
- The model picking the right tool vs. emitting malformed tool calls from too many similar schemas.
|
|
- The Chat Completions adapter staying inside structured-output limits vs. a 400 on tool-call payload size.
|
|
|
|
Removing them only shortens the direct tool list. The model still has `read`, `write`, `edit`, `exec`, `apply_patch`, image understanding, web search/fetch (when configured), memory, and session/agent tools. Extra catalogs stay reachable through Tool Search unless you set `tools.toolSearch: false`; explicit tool allows can opt a lean agent back into a trimmed workflow.
|
|
|
|
### When to turn it on
|
|
|
|
Enable lean mode once you have proved the model can talk to the Gateway but full agent turns misbehave:
|
|
|
|
1. `openclaw infer model run --gateway --model <ref> --prompt "Reply with exactly: pong"` succeeds.
|
|
2. A normal agent turn fails with malformed tool calls, oversized prompts, or the model ignoring its tools.
|
|
3. Toggling `localModelLean: true` clears the failure.
|
|
|
|
### When to leave it off
|
|
|
|
If your backend handles the full default runtime cleanly, leave this off. It is a workaround for local stacks that need a smaller tool surface, not a default for hosted models or well-resourced local rigs.
|
|
|
|
Lean mode does not replace `tools.profile`, `tools.allow`/`tools.deny`, or the model `compat.supportsTools: false` escape hatch. For a permanent narrower tool surface on a specific agent, prefer those stable knobs.
|
|
|
|
### Enable
|
|
|
|
```json5
|
|
{
|
|
agents: {
|
|
defaults: {
|
|
experimental: {
|
|
localModelLean: true,
|
|
},
|
|
},
|
|
},
|
|
}
|
|
```
|
|
|
|
For one agent only:
|
|
|
|
```json5
|
|
{
|
|
agents: {
|
|
entries: {
|
|
local: {
|
|
default: true,
|
|
model: "lmstudio/gemma-4-e4b-it",
|
|
experimental: {
|
|
localModelLean: true,
|
|
},
|
|
},
|
|
},
|
|
},
|
|
}
|
|
```
|
|
|
|
Restart the Gateway after changing the flag. Lean filtering removes `browser`, `cron`, `message`, `image_generate`, `music_generate`, `video_generate`, `tts`, and `pdf` unless you explicitly preserve them with `tools.allow` or `tools.alsoAllow`; Tool Search may still catalog preserved tools instead of exposing them directly.
|
|
|
|
## Experimental does not mean hidden
|
|
|
|
An experimental feature should say so plainly in docs and in the config path itself, not hide behind a stable-looking default knob.
|
|
|
|
## Related
|
|
|
|
- [Features](/concepts/features)
|
|
- [Release channels](/install/development-channels)
|