"Configure SwanLab settings, modes, credentials, hosts, login,
Scanned 9/9/2026
Install to Claude Code
npx -y skills add VectorSpaceLab/AREX-Skill --skill settings-and-modes --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Settings And Modes?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/vectorspacelab-settings-and-modes)More formats (shields.io, HTML) on the badges page.
---
name: settings-and-modes
description: "Configure SwanLab settings, modes, credentials, hosts, login,
require, and safe probe/terminal behavior."
disable-model-invocation: true
metadata:
disco-role: operating
license: Apache 2.0
---
# SwanLab Settings and Modes
Use this sub-skill when the task is about SwanLab configuration rather than experiment logging itself: `Settings`, `swanlab.merge_settings`, `swanlab.require`, run modes, login, credential storage, API/web hosts, environment/YAML/secret precedence, terminal capture, and environment or hardware probe boundaries.
## Route here for
- Choosing `online`, `offline`, `local`, or `disabled` mode for code, CLI defaults, CI, examples, or no-network environments.
- Constructing `swanlab.Settings(...)` or nested settings such as `Settings.Probe(...)`, `Settings.Terminal(...)`, `Settings.Run(...)`, and `Settings.Integration(...)`.
- Merging process-global settings before a run with `swanlab.merge_settings(...)`.
- Explaining `swanlab login`, `swanlab.login(...)`, `swanlab verify`, `swanlab logout`, local versus root credential saves, and non-interactive login.
- Configuring `api_key`, `api_host`, `web_host`, self-hosted/custom hosts, and host normalization.
- Diagnosing configuration source precedence across explicit kwargs, `swanlab.yaml`, `.env`, secrets, environment variables, per-directory config, user config, and netrc fallback.
- Disabling or bounding terminal proxying and probe collection for privacy, reproducibility, or optional hardware failures.
- Selecting transitional backend implementations with `swanlab.require(...)` or `SWANLAB_REQUIRE`.
## Route elsewhere
- Training-loop instrumentation, `swanlab.init`, `swanlab.log`, `finish`, scalar/config logging workflows, active run lifecycle, save policies, and async logging: `experiment-tracking`.
- Media objects, custom charts, image/audio/video/text/html/ECharts/molecule/3D details, and rich media optional dependencies: `media-and-custom-charts`.
- Uploading an offline/local run, validating run directories for sync, or converting W&B/TensorBoard/MLflow records: `sync-and-converters`.
- Object-oriented `swanlab.Api` calls and the `swanlab api ...` CLI surface: `open-api-and-cli`.
- Framework callbacks, notification plugins, CSV callbacks, and third-party integration imports: `integrations-and-plugins`.
## Read these references
- Settings fields, precedence, merge behavior, and probe/terminal controls: [references/configuration.md](references/configuration.md)
- Mode selection, login, credential storage, host handling, and self-hosted cautions: [references/modes-and-credentials.md](references/modes-and-credentials.md)
- Configuration and login failure diagnosis: [references/troubleshooting.md](references/troubleshooting.md)
## Safe local check
After SwanLab is installed, run the bundled checker to validate the local package behavior this sub-skill depends on:
```bash
python skills/disco/swanlab/sub-skills/settings-and-modes/scripts/check_settings_modes.py
```
The checker uses temporary directories, clears `SWANLAB_*` variables inside its process, performs disabled-mode and settings/merge assertions, and never uses credentials or network access.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!