weight-editing-harness
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| HARNESS_MODE | No | Selects the tool set and ops registered by the harness. Set to `llm` for transformers/MLPs via a ModelMap; the default mode is the Nav2D policy demo. | |
| HARNESS_BACKEND | No | Backend implementation to load, in `module:Class` form (e.g. `my_backend:MyBackend`), used when plugging the harness into your own model. The backend must implement `evaluate`, `forward`, `probe_obs`, and `load_ckpt`. | |
| HARNESS_RAW_VIEW | No | Set to `0` for a summary-only ablation arm of the `weights` tool; raw matrix slices (up to 4000 values) are returned otherwise. | |
| HARNESS_PYTHONPATH | No | Path to the directory containing your custom backend, e.g. `/path/to/your/backend`. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo recent community issues