Omnara

Create model provider config

model_providers_create

Create a model provider config, validate the configured credential where the provider exposes a validation endpoint, then probe the provider's /models endpoint. The result is returned as model_catalog; a failed probe does not fail creation.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "credential_secret_id": "<credential_secret_id>",
      "name": "<name>",
      "orgID": "<orgID>"
    },
    "name": "model_providers_create"
  }
}

Input Schema

Name	Required	Description
`name`	Yes	Human-readable name. Spaces and punctuation are allowed; leading or trailing whitespace and invisible or control characters are not.
`orgID`	Yes	Organization ID (org_...).
`preset`	No	Built-in provider settings: openai, openrouter, or anthropic. A preset fills in api_format, api_variant, base_url, endpoint_path, auth_kind, and auth_options, so omit those when using one.
`headers`	No	Extra HTTP headers sent with every request to this provider. Values are stored and returned as plain text; use secret_headers for keys and tokens.
`base_url`	No	Provider endpoint base URL.
`auth_kind`	No	How the credential is sent: bearer_token, api_key_header, or sigv4 (AWS).
`api_format`	No	Wire protocol the model endpoint speaks: openai-responses, openai-chat-completions, or anthropic-messages.
`api_variant`	No	Provider-specific flavor of the API format: default, openrouter, or bedrock.
`auth_options`	No	Non-secret authentication settings. Use an empty object with bearer_token. With api_key_header, set {"header_name":"..."}; Anthropic Messages defaults to {"header_name":"x-api-key"}. With sigv4, set the AWS signing service and region.
`endpoint_path`	No	URL path under base_url. Defaults to /responses for OpenAI Responses, /chat/completions for Chat Completions, and /messages for Anthropic Messages.
`secret_headers`	No	Extra HTTP headers whose values come from org-owned generic secrets, mapping header names to secret references.
`idle_timeout_ms`	No	Maximum wait for response headers or the next response bytes, in milliseconds. Heartbeats count as activity. Defaults to 300000 (5 minutes). Local response processing does not consume this idle budget.
`request_timeout_ms`	No	Total deadline for one provider attempt, including reasoning and response streaming, in milliseconds. Defaults to 3600000 (60 minutes). Retries have separate deadlines.
`credential_secret_id`	Yes	Secret ID (sec_...).

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedNov 14, 2026

changedInput schema / properties / base_url / description

Previous value: -"Provider endpoint base URL. Normal API mode requires a public HTTPS endpoint; localhost and loopback endpoints are accepted only when the API is running in insecure dev mode."New value: +"Provider endpoint base URL."

Changed7 schema fields changedOct 17, 2026

addedInput schema / $defs / SecretID / description
```
Added value: +"Secret ID (sec_...)."
```

addedInput schema / properties / api_format / description

Added value: +"Wire protocol the model endpoint speaks: openai-responses, openai-chat-completions, or anthropic-messages."

addedInput schema / properties / api_variant / description

Added value: +"Provider-specific flavor of the API format: default, openrouter, or bedrock."

addedInput schema / properties / auth_kind / description

Added value: +"How the credential is sent: bearer_token, api_key_header, or sigv4 (AWS)."

addedInput schema / properties / credential_secret_id / description
```
Added value: +"Secret ID (sec_...)."
```
addedInput schema / properties / orgID / description
```
Added value: +"Organization ID (org_...)."
```

addedInput schema / properties / preset / description

Added value: +"Built-in provider settings: openai, openrouter, or anthropic. A preset fills in api_format, api_variant, base_url, endpoint_path, auth_kind, and auth_options, so omit those when using one."

AddedOct 13, 2026

TDQS

A4.2/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare the safe-mutation profile (not read-only, not destructive, not idempotent); the description goes well beyond that by disclosing two side effects — credential validation where a validation endpoint exists, and a live /models probe — plus the key partial-failure rule that a failed probe does not fail creation. It also names the returned payload (model_catalog). One minor tension worth noting: the tool reaches out to a provider endpoint while openWorldHint is false, though the endpoint is tenant-configured so this is not a contradiction.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences, no filler, and the most important content (what is created, what side effects fire, what partial failure means) is front-loaded before any incidental detail.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 14-parameter nested-object mutation with no output schema, the description covers the operation and the return value well enough to call it correctly. It stops short of stating authorization requirements or whether a successful call guarantees the credential is usable, which would round out a complex creation tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and every field, including preset, auth_options, and the timeout parameters, is already documented in the schema. The description adds no parameter-level meaning such as preset-versus-manual exclusivity or which fields validation actually exercises, so the baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a precise verb+resource ("Create a model provider config") and immediately characterizes the full operation: credential validation, a /models probe, and the model_catalog result. That is enough to separate it from model_providers_get/list/update/delete without opening any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is only implied by the verb — an agent can infer it should call this when registering a new provider. There is no explicit when-to-use versus model_providers_update, no note that a credential secret must already exist, and no stated prerequisite permissions, so guidance is adequate but thin.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Omnara tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.