> For the complete documentation index, see [llms.txt](https://docs.algenta.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.algenta.ai/mcp-tools/capability-plane.md).

# Capability plane tools

MCP reference for Algenta's capability plane — list providers and bindings, discover and test them, route an objective, execute a capability, and manage skills.

The capability plane is Algenta's unified surface over every kind of capability — datasets, MCP tools and resources, prompt-skills, native tools, and runtime libraries — regardless of who provides them. These tools let an MCP client enumerate providers, create and test bindings, route an objective to the best capability with fallbacks, execute a capability, and manage skills. Each tool wraps a single authenticated `/v1/capabilit*` endpoint.

Configure `algenta-mcp` first — see [MCP server](/sdks/mcp.md). For the concepts behind providers, bindings, and execution ownership, see [The capability plane](/concepts/capability-plane.md).

{% hint style="info" %}
The core loop is **route → execute**: `route_capabilities` returns the best capability and its authoritative `execution_owner`, and `execute_capability` runs it. Only `algenta_managed` routes execute server-side; `client_managed` routes must run in your own app or adapter.
{% endhint %}

## Providers and bindings

| Tool                          | What it does                                                                          | Key inputs                                                                                                                                                           |
| ----------------------------- | ------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `list_capability_providers`   | List unified providers across data, MCP, skills, native tools, and runtime libraries. | *(none)*                                                                                                                                                             |
| `list_capability_bindings`    | List bindings for the current organization.                                           | `provider_id`, `scope` (`user`\|`workspace`\|`organization`)                                                                                                         |
| `create_capability_binding`   | Create one binding for a provider/profile pair.                                       | `provider_id`, `profile_id`, `binding_name` (required); `scope`, `scope_ref`, `execution_owner` (`algenta_managed`\|`client_managed`), `config`, `customer_metadata` |
| `test_capability_binding`     | Test a saved binding, or preview-test an unsaved one.                                 | `binding_id` (saved) **or** `provider_id`+`profile_id`+`config` (preview); `scope`, `scope_ref`, `execution_owner`, `customer_metadata`                              |
| `discover_capability_binding` | Discover the capabilities a binding exposes, or preview-discover an unsaved one.      | `binding_id` (saved) **or** `provider_id`+`profile_id`+`config` (preview); `scope`, `scope_ref`, `execution_owner`, `customer_metadata`                              |

## Capabilities

| Tool                 | What it does                                                                                     | Key inputs                                                                                                                                |
| -------------------- | ------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------- |
| `list_capabilities`  | List unified capabilities, filtered by kind, provider, or binding.                               | `kinds` (`dataset`, `mcp_tool`, `mcp_resource`, `mcp_prompt`, `skill`, `native_tool`, `runtime_library`), `provider_ids`, `binding_ids`   |
| `get_capability`     | Get one capability by id.                                                                        | `capability_id` (required); `include_instruction`                                                                                         |
| `route_capabilities` | Route an objective to the best capability with fallbacks and an authoritative `execution_owner`. | `objective` (required); `binding_ids`, `provider_ids`, `kinds`, `execution_owners`, `artifact_affinities`, `tags`, `max_fallbacks` (0–10) |
| `execute_capability` | Execute one routed or known `algenta_managed` capability by id.                                  | `capability_id` (required); `binding_id`, `input`, `request_id`                                                                           |

## Skills

| Tool            | What it does                                                 | Key inputs                                                                                              |
| --------------- | ------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------- |
| `list_skills`   | List skill capabilities from the plane.                      | *(none)*                                                                                                |
| `enable_skill`  | Enable one prompt-skill as a first-class capability binding. | `skill_name`, `instruction` (required); `description`, `tags`, `artifact_affinities`, `execution_owner` |
| `disable_skill` | Disable one skill binding by id.                             | `binding_id` (required)                                                                                 |

## Endpoint mapping

| Tool                          | Method   | Endpoint                                                                              |
| ----------------------------- | -------- | ------------------------------------------------------------------------------------- |
| `list_capability_providers`   | `GET`    | `/v1/capability-providers`                                                            |
| `list_capability_bindings`    | `GET`    | `/v1/capability-bindings`                                                             |
| `create_capability_binding`   | `POST`   | `/v1/capability-bindings`                                                             |
| `test_capability_binding`     | `POST`   | `/v1/capability-bindings/{binding_id}/test` or `/v1/capability-bindings/test`         |
| `discover_capability_binding` | `POST`   | `/v1/capability-bindings/{binding_id}/discover` or `/v1/capability-bindings/discover` |
| `list_capabilities`           | `GET`    | `/v1/capabilities`                                                                    |
| `get_capability`              | `GET`    | `/v1/capabilities/{capability_id}`                                                    |
| `route_capabilities`          | `POST`   | `/v1/capabilities/route`                                                              |
| `execute_capability`          | `POST`   | `/v1/capabilities/execute`                                                            |
| `list_skills`                 | `GET`    | `/v1/capabilities?kinds=skill`                                                        |
| `enable_skill`                | `POST`   | `/v1/capability-bindings` (then discover)                                             |
| `disable_skill`               | `DELETE` | `/v1/capability-bindings/{binding_id}`                                                |

## Worked examples

### Enumerate what is available

Start with the providers on the instance, then narrow to concrete capabilities by kind.

{% code title="list\_capability\_providers (tool call)" %}

```json
{ "name": "list_capability_providers", "arguments": {} }
```

{% endcode %}

{% code title="list\_capabilities (tool call)" %}

```json
{
  "name": "list_capabilities",
  "arguments": {
    "kinds": ["mcp_tool", "dataset"]
  }
}
```

{% endcode %}

```bash
curl -sS "$ALGENTA_BASE_URL/v1/capabilities?kinds=mcp_tool&kinds=dataset" \
  -H "Authorization: Bearer $ALGENTA_API_KEY" | jq '.capabilities[].capability_id'
```

### Route an objective, then execute it

`route_capabilities` returns a ranked capability plus fallbacks and the authoritative `execution_owner`. If that owner is `algenta_managed`, run it with `execute_capability`.

{% code title="route\_capabilities (tool call)" %}

```json
{
  "name": "route_capabilities",
  "arguments": {
    "objective": "summarize the latest churn report",
    "kinds": ["skill", "mcp_tool"],
    "max_fallbacks": 2
  }
}
```

{% endcode %}

{% code title="execute\_capability (tool call)" %}

```json
{
  "name": "execute_capability",
  "arguments": {
    "capability_id": "$CAPABILITY_ID",
    "input": { "period": "2026-Q2" }
  }
}
```

{% endcode %}

```bash
curl -sS "$ALGENTA_BASE_URL/v1/capabilities/execute" \
  -H "Authorization: Bearer $ALGENTA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "capability_id": "$CAPABILITY_ID", "input": { "period": "2026-Q2" } }'
```

{% hint style="warning" %}
`execute_capability` runs only `algenta_managed` routes server-side. A `client_managed` route is returned by `route_capabilities` for you to execute in your own app or adapter path — the engine will not run it on your behalf.
{% endhint %}

### Enable a prompt-skill

Register an instruction-only skill as a first-class capability binding, then let routing pick it up.

{% code title="enable\_skill (tool call)" %}

```json
{
  "name": "enable_skill",
  "arguments": {
    "skill_name": "Executive summary",
    "instruction": "Summarize the input into five crisp bullet points for an executive audience.",
    "tags": ["summarization", "reporting"]
  }
}
```

{% endcode %}

Disable it later with `disable_skill` and the returned `$BINDING_ID`.

## Related pages

{% content-ref url="/pages/bjkEhVUT4VwGkNpsxVeo" %}
[The capability plane](/concepts/capability-plane.md)
{% endcontent-ref %}

{% content-ref url="/pages/2P0FmT4EqEhuumZOxjAJ" %}
[Route & execute capabilities](/guides/capabilities.md)
{% endcontent-ref %}

{% content-ref url="/pages/dbMexAzWYilfd77tn3O8" %}
[API endpoint reference](/http-api/reference.md)
{% endcontent-ref %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.algenta.ai/mcp-tools/capability-plane.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.