Numonic

npx add-mcp 'https://www.numonic.ai/mcp'

MCP Server Reference

Connect AI agents to your digital asset library via Model Context Protocol. 66 tools across 13 domains, 5 resources, and 5 guided prompts — ready for Claude, ChatGPT, Gemini, Codex, VS Code, Cursor, and custom agents built on the MCP SDKs. Every MCP tool has a REST equivalent in the REST API Reference.

Tools

66

Resources

5

Prompts

5

Transport

HTTP / JSON-RPC 2.0

Endpoint: POST https://www.numonic.ai/mcp

Getting Started

New here? The connect-an-agent quickstart walks through both ways to connect — an API key, or OAuth for clients that only take a connector URL.

1. 1

   Create an API key

   In the Numonic dashboard, go to Settings → Workspace → Connected Agents and create a new key. Keys are available on all tiers, including free. Your key looks like napi_a1b2c3d4e5f6… — store it securely; it won't be shown again.
2. 2

   Connect your MCP client

   The Numonic MCP server accepts connections over Streamable HTTP (JSON-RPC 2.0 over HTTP POST). See Client Configuration for how to connect Claude, ChatGPT, Gemini, Codex, VS Code, Cursor, and other clients.

3. Test the connection

   Send an initialize request to verify everything works:

   curl -X POST https://www.numonic.ai/mcp \
     -H "X-API-Key: napi_your_key_here" \
     -H "Content-Type: application/json" \
     -d '{
       "jsonrpc": "2.0",
       "method": "initialize",
       "params": { "protocolVersion": "2025-03-26" },
       "id": 1
     }'
   

   A successful response returns the server capabilities. Then list tools with {"method":"tools/list"}.

Client Configuration

Chat apps connect with OAuth: you paste the URL and sign in to Numonic. Developer tools take an API key. After editing a config file, restart the client.

Apps: paste the URL and sign in

https://www.numonic.ai/mcp

You pick the workspace and permissions on the Numonic consent page. OAuth connection steps

Claude (desktop app and claude.ai)

Settings → Connectors → Add custom connector. Paste the URL, then sign in to Numonic and approve access.

Do not add Numonic to claude_desktop_config.json: that file only starts local servers. Free plans allow one custom connector.

ChatGPT

Settings → Security and login → turn on Developer mode. Then open Plugins, select +, and create an app with the URL.

Plus, Pro, Business, Enterprise and Education plans, on the web.

Gemini app

gemini.google.com → Settings → Connected Apps → Add a custom app. Enter the URL, select Next, and sign in.

Personal Google accounts only (not work or school), aged 18+ in the US, in English, with Keep Activity on.

Grok

grok.com/connectors → New Connector → Custom. Enter the URL and sign in.

Le Chat (Mistral)

Connectors → + Add Connector → Custom MCP Connector. Name it numonic, enter the URL, select Connect, and sign in.

Admins only; on Free, Pro and Student plans the account owner is the admin.

Perplexity

Add a custom remote connector with the URL and choose OAuth. API key authentication with your napi_ key also works.

Pro and Enterprise plans.

Microsoft Copilot Studio

Your agent's Tools → Add a tool → New tool → Model Context Protocol. Enter the URL, then choose OAuth 2.0 → Dynamic discovery.

Or choose API key → Header, with header name X-API-Key, and supply your napi_ key.

Developer tools: use an API key

Claude Code (CLI)

Run this command in your terminal.

claude mcp add --transport http numonic https://www.numonic.ai/mcp \
  --header "Authorization: Bearer napi_your_key_here"

Codex (CLI and IDE extension)

Codex reads the key from an environment variable, so set it where Codex runs. The server is saved to ~/.codex/config.toml, which the IDE extension also reads. To sign in with OAuth instead, leave out --bearer-token-env-var and run codex mcp login numonic.

export NUMONIC_API_KEY=napi_your_key_here
codex mcp add numonic --url https://www.numonic.ai/mcp \
  --bearer-token-env-var NUMONIC_API_KEY

Gemini CLI

Run this command, or add the server by hand to ~/.gemini/settings.json (.gemini/settings.json for one project) under mcpServers. By hand, use httpUrl, not url: Gemini CLI reads url as an SSE endpoint.

gemini mcp add --transport http numonic https://www.numonic.ai/mcp \
  --header "Authorization: Bearer napi_your_key_here"

VS Code (GitHub Copilot)

Save as.vscode/mcp.json in your workspace, or add it to your user MCP configuration. VS Code asks for the key when the server starts. Visual Studio 2022 (17.14+) reads servers from.mcp.json in the solution folder or %USERPROFILE%\.mcp.json; there, list only the url and choose Manage Authentication to sign in with OAuth.

{
  "inputs": [
    {
      "type": "promptString",
      "id": "numonic-key",
      "description": "Numonic API key",
      "password": true
    }
  ],
  "servers": {
    "numonic": {
      "type": "http",
      "url": "https://www.numonic.ai/mcp",
      "headers": {
        "Authorization": "Bearer ${input:numonic-key}"
      }
    }
  }
}

Cursor / Devin Desktop (formerly Windsurf)

Cursor: add to.cursor/mcp.json in your project, or ~/.cursor/mcp.json for every project. Devin Desktop: add to ~/.config/devin/mcp_config.json (%APPDATA%\devin\mcp_config.json on Windows).

{
  "mcpServers": {
    "numonic": {
      "url": "https://www.numonic.ai/mcp",
      "headers": {
        "Authorization": "Bearer napi_your_key_here"
      }
    }
  }
}

Custom Agents (SDKs)

TypeScript

@modelcontextprotocol/sdk ↗

import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StreamableHTTPClientTransport }
  from '@modelcontextprotocol/sdk/client/streamableHttp.js';

const transport = new StreamableHTTPClientTransport(
  new URL('https://www.numonic.ai/mcp'),
  {
    requestInit: {
      headers: { Authorization: 'Bearer napi_your_key_here' },
    },
  }
);

const client = new Client({ name: 'my-agent', version: '1.0.0' });
await client.connect(transport);

// Search for assets
const result = await client.callTool('SearchAssets', {
  query: 'tool:midjourney AND tag:approved',
  limit: 10,
});
console.log(result.content);

Python

mcp ↗

from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client

async with streamablehttp_client(
    "https://www.numonic.ai/mcp",
    headers={"Authorization": "Bearer napi_your_key_here"},
) as (read_stream, write_stream, _):
    async with ClientSession(read_stream, write_stream) as session:
        await session.initialize()

        tools = await session.list_tools()
        print(f"{len(tools.tools)} tools available")

        result = await session.call_tool(
            "SearchAssets",
            arguments={
                "query": "tool:midjourney AND tag:approved",
                "limit": 10,
            },
        )
        print(result.content)

Authentication

Which credential to use

Client	Credential	Notes
Connector apps: Claude, ChatGPT, Gemini, Grok, Le Chat, Perplexity, Copilot Studio	OAuth 2.1 (authorization code + PKCE)	The client discovers the authorization server from the endpoint, registers itself, and sends you to a Numonic consent page. You choose the workspace and permissions there. No key to paste.
Claude Code, Codex, Gemini CLI, VS Code, Cursor, Devin Desktop (formerly Windsurf), scripts and custom agents	API key (napi_...) in a header	Create the key under Settings → Workspace → Connected Agents and send it as shown in Client Configuration.
Anything using a Numonic or Supabase login session token	Not accepted	A session or access token from signing in to Numonic is rejected with 401. Use an API key, or connect through OAuth.

OAuth connection permissions

On the consent page you grant a connection some of the following Numonic permissions. They are separate from the OAuth scopes the client requests (identity only, such as openid). A connection acts as its own agent, not as you, and you can disconnect it at any time under Settings → Workspace → OAuth connectors. Reconnecting a connector replaces its earlier connection.

read	View assets, collections and their details
search	Search your library
write	Add and update assets, tags and collections
export	Export and publish assets

admin (running pipelines, registering webhooks) cannot be granted to an OAuth connection. Use a dedicated API key for that.

Rolling out: OAuth connections can register and sign in today, but tool calls from an OAuth connection are refused until per-connection permissions ship. Until then, use an API key for tool access.

API keys

All API keys must start with the napi_ prefix. Keys are SHA-256 hashed server-side — Numonic never stores your raw key.

Method	Header	Note
Bearer token	Authorization: Bearer napi_...	Required by most MCP clients
API key header	X-API-Key: napi_...	Preferred for direct HTTP
Legacy header	api-key: napi_...	Backwards compatibility

Multi-tenant access

If your API key has access to multiple tenants, pass X-Tenant-ID to select which tenant to operate on. If omitted, the key's default tenant is used.

Tools(66)

Atomic operations your agent can call via tools/call. Each tool accepts a JSON arguments object and returns structured results.

Asset Ingestion1 tool

Store and ingest assets into Numonic

StoreAsset

tool · asset ingestion

Store an asset (file) in Numonic with metadata and lineage. Provide EXACTLY ONE of: base64 bytes (asset_data_base64), a reference to a pre-uploaded object (asset_storage_ref), or an HTTPS URL the Edge Function will fetch server-side (asset_data_url). The URL channel accepts allowlisted hosts only (configured per-deployment via MCP_URL_INGEST_ALLOWLIST; default: signed Supabase URLs + Comfy Cloud + Midjourney CDN + Civitai + Numonic-managed S3). Default size cap 100 MB (MCP_URL_INGEST_MAX_BYTES); default fetch timeout 30s (override per-call via timeout_seconds, max 300). Distinct error codes: URL_EXPIRED, URL_FORBIDDEN_HOST, URL_TOO_LARGE, URL_FETCH_TIMEOUT, URL_CONTENT_TYPE_MISMATCH. Lineage is determined by the server: after upload Numonic extracts the prompt, workflow and models embedded in the file (e.g. ComfyUI PNG/MP4 metadata) and GetAssetDetails reports them with provenance.lineage_source. prompt_metadata is OPTIONAL — omit it when you do not know how the asset was made (never invent a prompt or agent hash); an asserted empty or "none" value never overrides what extraction finds. IMPORTANT for chat-attached files: developer-mode MCP connectors do not reliably receive a file reference for a file attached directly in chat (OpenAI Apps SDK openai/fileParams hydration is documented for first-party Apps SDK directory apps and is unconfirmed for custom developer-mode connectors — see GitLab #2807). If the user attaches an image/file in chat and asks you to store it, do NOT re-encode a large attachment as asset_data_base64 yourself — a large inline re-encode can run for minutes and get cancelled. Ask the user for a URL to pass via asset_data_url instead, or use asset_storage_ref if the file was already uploaded to a Numonic-issued signed URL.

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	Tenant ID (optional — auto-injected from API key auth, omit for automatic tenant detection)
filename	string	required	Original filename (non-empty)
mime_type	string	required	MIME type of the asset (e.g., 'image/png', 'text/plain')
asset_data_base64	string	optional	Base64-encoded asset bytes for small assets. Provide either this OR asset_storage_ref.
asset_data_url	string ·uri	optional	HTTPS URL the Edge Function will fetch server-side. Allowlisted hosts only. Mutually exclusive with asset_data_base64 / asset_storage_ref. Returns dedicated error codes (URL_EXPIRED / URL_FORBIDDEN_HOST / URL_TOO_LARGE / URL_FETCH_TIMEOUT / URL_CONTENT_TYPE_MISMATCH).
allow_content_type_mismatch	boolean	optional	When asset_data_url is used, accept responses whose Content-Type does not match mime_type. Defaults to true because Comfy Cloud serves PNGs as application/octet-stream.
timeout_seconds	integer	optional	Per-call fetch timeout for asset_data_url (5–300s). Default 30.
asset_storage_ref	object	optional	Reference to asset data already uploaded via signed URL. Mutually exclusive with asset_data_base64 / asset_data_url.
prompt_metadata	object	optional	OPTIONAL client assertion of how this asset was generated (prompt + responding agent). Omit it if you do not know; the server extracts lineage from the stored file and extracted data always wins over an empty or "none" assertion.
asset_metadata	object	optional	Optional descriptive metadata for the asset.
tags	string[]	optional	Optional array of tag names to apply to the asset.
collection_h	string	optional	Optional Collection Hash (SHA-1) to add this asset to immediately on creation.
parent_asset_h	string	optional	Optional Asset Hash (SHA-1) of a parent asset if this is a version or derivation.

Asset Discovery3 tools

Retrieve individual assets, public URLs, and creative sessions

GetMidjourneyEvolutionChain

tool · asset discovery

Retrieve the evolution chain (parent→child lineage) for a Midjourney asset, showing how it evolved through variations and upscales.

Parameter	Type	Required	Description
asset_id	string	required	The Asset Hash (SHA-1) of the asset to get evolution chain for
max_depth	number	optional	Maximum depth to traverse (default: 10, max: 50)

GetCreativeSession

tool · asset discovery

Discover all Midjourney assets created within a time window of a given asset, grouped by temporal proximity to form a "creative session".

Parameter	Type	Required	Description
asset_id	string	required	The Asset Hash (SHA-1) of the asset to find session for
time_window	string	optional	Time window for session discovery (e.g., "2 hours", "30 minutes", "1 day")

GetAssetPublicUrl

tool · asset discovery

Reverse lookup: finds if an asset is in ANY published collection and returns the public URL. Returns one of three states: (1) Asset IS published with public_url, collection_path, preset, published_at; (2) Asset NOT published but in collections: lists collection paths; (3) Asset NOT in any collection: empty collections array.

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	Tenant UUID (optional - auto-injected from API key auth)
asset_h	string	required	Asset hash (SHA-1, 40 hex chars) to look up

Annotations4 tools

Create, read, update, and delete asset annotations

CreateAnnotation

tool · annotations

Create a new annotation on a ComfyUI workflow node with full audit trail.

Parameter	Type	Required	Description
workflow_node_id	string	required	40-character hexadecimal workflow node identifier
content	string	required	Annotation text content (1-10000 characters)
content_type	string	optional	Content format type enum: text · markdown · json
visibility	string	optional	Visibility level: private (author only), team (tenant), public (all) enum: private · team · public

GetAnnotations

tool · annotations

Retrieve all annotations for a specific ComfyUI workflow node with author information.

Parameter	Type	Required	Description
workflow_node_id	string	required	40-character hexadecimal workflow node identifier
include_resolved	boolean	optional	Include resolved annotations in results (default: false)

UpdateAnnotation

tool · annotations

Update an existing annotation by creating a new version via supersession chain (effectivity pattern).

Parameter	Type	Required	Description
annotation_id	string	required	40-character hexadecimal annotation identifier
content	string	optional	Updated annotation content (optional)
visibility	string	optional	Updated visibility level (optional) enum: private · team · public

DeleteAnnotation

tool · annotations

Delete (soft delete) an annotation - marks it as deleted without removing data.

Parameter	Type	Required	Description
annotation_id	string	required	40-character hexadecimal annotation identifier

Publishing3 tools

Publish collections and retrieve public-access URLs

PublishCollection

tool · publishing

Publish a collection with specified privacy and publish presets, making assets publicly accessible. Applies ADR-057 metadata stripping rules and generates public URLs.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
collection_h	string	required	Collection hash (SHA-1, 40 hex chars) to publish
privacy_preset	string	optional	Privacy preset for metadata stripping (ADR-057): share (strip workflow/models/GPS), portfolio (default, keeps attribution), client (business delivery), archive (keep all metadata) enum: share · portfolio · client · archive
publish_preset	string	optional	Image optimization preset: web-standard (default), high-quality, thumbnail enum: web-standard · high-quality · thumbnail

UnpublishCollection

tool · publishing

Remove a collection from public access. Published assets are marked as unpublished and deleted from public storage.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
collection_h	string	required	Collection hash (SHA-1, 40 hex chars) to unpublish

GetCollectionPublicUrls

tool · publishing

Get publication status and public URLs for all assets in a collection. Returns whether the collection is published, publication metadata, and URLs for each published asset.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
collection_h	string	required	Collection hash (SHA-1, 40 hex chars) to get public URLs for

Export2 tools

Export assets using configurable presets

ExportAssets

tool · export

Not available yet: file export over MCP produces no file, so this tool validates its input and returns an EXPORT_UNAVAILABLE error. To share images, use PublishCollection, which returns a public link to signed copies; to download files, use the Numonic app or the REST endpoint POST /api/v1/assets/export.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the target tenant for this export operation
asset_hs	string[]	required	Array of Asset Hashes (SHA1) to be exported
preset	string	optional	Export preset: share (social media, max privacy), portfolio (keeps attribution), client (business delivery), archive (full metadata), custom (user-defined) enum: share · portfolio · client · archive · custom
options	object	optional	Custom options (only when preset is "custom"). Fine-grained control over metadata stripping.
format	string	optional	Output format (preserve = keep source format) enum: png · jpeg · webp · preserve
export_configuration_h	string	optional	Legacy: Export Configuration Hash. Mutually exclusive with preset.

ListExportPresets

tool · export

List available export presets with their default options. Returns the preset configurations used by privacy-aware exports in the Numonic app and REST API (ADR-057).

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	Tenant UUID (for tenant-specific presets if available)

Analytics2 tools

Query search analytics and tenant storage metrics

GetTenantStorageDetails

tool · analytics

Retrieve storage usage details for a tenant including bytes used, GB used, storage limit, percentage used, and over-limit status. Helps monitor storage consumption and enforce storage quotas.

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	Optional: UUID of the tenant to query. If omitted, uses the tenant_id from auth context (current user's tenant).

GetSearchAnalytics

tool · analytics

Retrieve search analytics summary with health assessment (green/amber/red) for monitoring search quality. Returns zero-result rate, latency percentiles (P50/P95/P99), query-type breakdown, and top zero-result queries. Health thresholds: zero-result rate (<15% green, 15-25% amber, >25% red), P95 latency (<500ms green, 500-1000ms amber, >1000ms red).

Parameter	Type	Required	Description
period	integer	optional	Number of days to aggregate (1-90). Default: 7.
tenant_id	string ·uuid	optional	Optional: UUID of the tenant to query. If omitted, returns analytics for all tenants or current user's tenant.

Pipelines5 tools

Execute, save, list, and run reusable processing pipelines

ExecutePipeline

tool · pipelines

Execute a multi-stage asset pipeline. Compose select, filter, transform, action, output, and summarize stages into a single operation. Use dry_run: true to preview changes before committing. Stages by category: SELECT: search, collection, ids, diff (set comparison). FILTER: where, sort, limit, deduplicate, sample. TRANSFORM: set_tag, remove_tag, set_field, regex_replace, compute, set_visibility, set_owner, enrich (stub), approve (stub). ACTION: add_to_collection, move, delete, archive. OUTPUT: export (ADR-057 privacy-aware), notify (stub), tee (passthrough fan-out). SUMMARIZE: count, group_by, stats, histogram. Output stages require confirm: true and are skipped during dry-run. Summarize stages execute during dry-run to provide aggregation previews.

Parameter	Type	Required	Description
stages	object[]	required	Ordered list of pipeline stages. Must start with a select stage (search, collection, ids, or diff).
dry_run	boolean	optional	Preview what would happen without executing mutations. Transform/action stages show simulated side effects. Output stages are skipped. Summarize stages execute to provide aggregation previews. Default: false.
timeout_ms	number	optional	Maximum execution time in milliseconds (1000-60000). Default: 30000.

SavePipeline

tool · pipelines

Create or update a named saved pipeline. Provide pipeline_definition_h to update an existing pipeline (SCD Type 2 versioning preserves full edit history). Omit it to create a new pipeline. Pipelines are identified by slug (derived from name) and stored with full stage definitions.

Parameter	Type	Required	Description
name	string	required	Pipeline name (used to generate URL-safe slug). Must be unique per tenant.
stages	object[]	required	Ordered pipeline stage definitions (same format as ExecutePipeline stages)
description	string	optional	Human-readable description of what this pipeline does
default_dry_run	boolean	optional	Default dry_run setting when run without override (default: false)
default_timeout_ms	number	optional	Default timeout in ms, 1000-60000 (default: 30000)
tags	string[]	optional	Labels for organizing pipelines (e.g., ["weekly", "client-delivery"])
pipeline_definition_h	string	optional	40-char hex hash of existing pipeline to update. Omit to create new.

ListPipelines

tool · pipelines

List saved pipelines for the current tenant. Returns pipeline metadata including name, stage count, tags, and default execution settings. Use pipeline_definition_h from results with RunSavedPipeline to execute.

Parameter	Type	Required	Description
tags	string[]	optional	Filter by tags (e.g., ["weekly"])

RunSavedPipeline

tool · pipelines

Execute a saved pipeline by ID or name. Supports overriding default_dry_run and default_timeout_ms at run time. Returns the same execution trace as ExecutePipeline. Provide either pipeline_definition_h (40-char hex) or pipeline_name (resolved automatically).

Parameter	Type	Required	Description
pipeline_definition_h	string	optional	40-char hex hash of the pipeline to run
pipeline_name	string	optional	Pipeline name (alternative to pipeline_definition_h). Resolved to hash automatically.
dry_run	boolean	optional	Override the saved default_dry_run. Defaults to pipeline's saved setting.
timeout_ms	number	optional	Override the saved default_timeout_ms (1000-60000)
vars	object	optional	Runtime variable injection (merged into pipeline context vars)

ListPipelineTemplates

tool · pipelines

List available pipeline templates for the current tenant. Returns system templates (available to all tenants) plus any tenant-specific templates. Use the pipeline_definition_h from results with POST /api/v1/pipelines/templates to clone a template into a new saved pipeline.

Pipeline Stages6 tools

Execute individual pipeline stages (select, filter, transform, etc.)

SelectStage

tool · pipeline stages

Execute a single SELECT-category pipeline stage. SELECT stages are entry points that produce an initial asset set. Supported stage_type values: search (text query), collection (load from collection path), ids (explicit asset hashes), diff (set difference between two sub-selects).

Parameter	Type	Required	Description
stage_type	string	required	The SELECT stage type to execute enum: search · collection · ids · diff
config	object	required	Stage-specific configuration. search: { query: "tag:approved" }. collection: { path: "projects.nike", include_nested: true }. ids: { asset_ids: ["abc..."] }. diff: { set_a: {...}, set_b: {...}, mode: "only_in_a" }.
dry_run	boolean	optional	If true, simulates execution without side effects

FilterStage

tool · pipeline stages

Execute a single FILTER-category pipeline stage on a set of assets. Supported stage_type values: where (field filter), sort (order by field), limit (cap result count), deduplicate (remove duplicates), sample (random subset).

Parameter	Type	Required	Description
stage_type	string	required	The FILTER stage type to execute enum: where · sort · limit · deduplicate · sample
config	object	optional	Stage-specific configuration. sort: { by: "created_at", order: "desc" }. limit: { count: 20 }. where: { field: "tool", operator: "eq", value: "midjourney" }.
assets	string[]	required	Array of asset hashes to filter. Required for non-SELECT stages.
dry_run	boolean	optional	If true, simulates execution without side effects

TransformStage

tool · pipeline stages

Execute a single TRANSFORM-category pipeline stage on a set of assets. Modifies asset metadata or triggers AI enrichment. Supported stage_type values: set_tag, remove_tag, set_field, regex_replace, compute, set_visibility, set_owner, enrich, approve.

Parameter	Type	Required	Description
stage_type	string	required	The TRANSFORM stage type to execute enum: set_tag · remove_tag · set_field · regex_replace · compute · set_visibility · set_owner · enrich · approve
config	object	optional	Stage-specific configuration. set_tag: { tags: ["approved"] }. enrich: { operations: ["auto_tag"] }. set_field: { field: "status", value: "reviewed" }.
assets	string[]	required	Array of asset hashes to transform
dry_run	boolean	optional	If true, simulates execution without side effects

ActionStage

tool · pipeline stages

Execute a single ACTION-category pipeline stage on a set of assets. Performs structural operations like moving, deleting, or archiving assets. Supported stage_type values: add_to_collection, move, delete, archive.

Parameter	Type	Required	Description
stage_type	string	required	The ACTION stage type to execute enum: add_to_collection · move · delete · archive
config	object	optional	Stage-specific configuration. add_to_collection: { path: "projects.nike" }. move: { from: "inbox", to: "approved" }. delete: { confirm: true }. archive: { confirm: true }.
assets	string[]	required	Array of asset hashes to act on
dry_run	boolean	optional	If true, simulates execution without side effects

OutputStage

tool · pipeline stages

Execute a single OUTPUT-category pipeline stage on a set of assets. Produces outputs like exports, notifications, or tee copies. Supported stage_type values: export, notify, tee.

Parameter	Type	Required	Description
stage_type	string	required	The OUTPUT stage type to execute enum: export · notify · tee
config	object	optional	Stage-specific configuration. export: { preset: "client", confirm: true }. notify: { event_type: "pipeline.completed" }. tee: { action: { type: "add_to_collection", path: "backup" } }.
assets	string[]	required	Array of asset hashes for output
dry_run	boolean	optional	If true, simulates execution without side effects

SummarizeStage

tool · pipeline stages

Execute a single SUMMARIZE-category pipeline stage on a set of assets. Produces aggregate data and statistics. Supported stage_type values: count, group_by, stats, histogram.

Parameter	Type	Required	Description
stage_type	string	required	The SUMMARIZE stage type to execute enum: count · group_by · stats · histogram
config	object	optional	Stage-specific configuration. count: { group_by: "tool" }. group_by: { field: "mime_type" }. stats: { field: "file_size" }. histogram: { field: "created_at" }.
assets	string[]	required	Array of asset hashes to summarize
dry_run	boolean	optional	If true, simulates execution without side effects

Automation Rules3 tools

Create, list, and trigger automation rules

CreateRule

tool · automation rules

Create an automation rule that triggers a saved pipeline on events, schedules, or watch intervals. Provide pipeline_definition_h (40-char hex) or pipeline_name (resolved automatically). Trigger types: event (fires on asset/pipeline events), schedule (cron-based), watch (periodic query check).

Parameter	Type	Required	Description
name	string	required	Rule name (unique per tenant)
description	string	optional	Human-readable description of what this rule does
trigger_config	object	required	Trigger configuration with type discriminator. Event: { type: "event", event_type: "pipeline.completed" }. Schedule: { type: "schedule", cron: "0 9 * * 1" }. Watch: { type: "watch", query: "tag:unreviewed", interval_minutes: 60 }.
rate_limit_config	object	optional	Optional rate limiting config (e.g., { max_executions_per_hour: 10 })
pipeline_vars	object	optional	Variables to inject into pipeline context when rule fires
pipeline_definition_h	string	optional	40-char hex hash of the saved pipeline to trigger
pipeline_name	string	optional	Pipeline name (alternative to pipeline_definition_h). Resolved to hash automatically.

ListRules

tool · automation rules

List automation rules for the current tenant with optional filters. Returns rule metadata including name, trigger type, enabled status, and linked pipeline.

Parameter	Type	Required	Description
enabled	boolean	optional	Filter by enabled/disabled status
trigger_type	string	optional	Filter by trigger type enum: event · schedule · watch
tenant_id	string ·uuid	optional	Tenant UUID (optional - auto-injected from API key auth)

TriggerRule

tool · automation rules

Manually trigger automation rule processing. For schedule/watch: evaluates all due rules for the tenant. For event: dispatches to matching event-based rules. Returns execution results for each triggered rule.

Parameter	Type	Required	Description
trigger_type	string	required	Type of trigger to process: schedule (check cron-due rules), watch (check query-based rules), event (dispatch event to matching rules) enum: schedule · watch · event
event_type	string	optional	Event type to dispatch (required when trigger_type is "event", e.g., "pipeline.completed")
event_data	object	optional	Event payload data (optional, passed to matching event rules)
tenant_id	string ·uuid	required	Tenant UUID (required for event dispatch)

Webhooks1 tool

Register webhooks for event-driven integrations

RegisterWebhook

tool · webhooks

Register a new webhook subscription to receive event notifications. Returns webhook ID and signing secret (shown once). Supports events: pipeline.completed, pipeline.failed, pipeline.export.completed, test.ping.

Parameter	Type	Required	Description
name	string	required	Human-readable name for the webhook subscription
url	string	required	HTTPS endpoint URL to receive webhook events (must be public, no private IPs)
events	string[]	required	Event types to subscribe to (e.g., ["pipeline.completed", "pipeline.failed"])
description	string	optional	Optional description of the webhook purpose
enabled	boolean	optional	Whether the webhook is enabled (default: true)
timeout_ms	number	optional	Request timeout in milliseconds, 1000-60000 (default: 10000)

Uncategorised29 tools

Uncategorised

GetAssetDetails

tool · uncategorised

Retrieve comprehensive details for a specific asset by its SHA-1 hash, including how it was made. Answers 'which prompt, model and workflow produced this?' directly — no download needed: `prompts.positive`/`prompts.negative` hold the prompt, `models[]` lists every model the workflow loaded (name, type such as checkpoint/unet/lora/vae, loader node_id), `workflow.workflow_json` is the full ComfyUI graph (with `workflow.tool_name`, `node_count`, `custom_nodes`), and `generation_parameters` holds sampler settings (model_name, seed, steps, cfg_scale, sampler, scheduler) when they were read from a standard KSampler node, else null (the graph still has them). `provenance.lineage_source` says where that lineage came from: 'extracted' (read by Numonic from the stored file; `prompts.source`/`workflow.source` are then 'extracted'), 'client_asserted' (only the uploader's claim), or 'unknown' (nothing asserted or found). Also returns title, description, tags, collections, provenance (creation prompt, parent assets), embedding info and copyright. For Model Library records of those models (base model, trigger words, usage) call GetAssetModels. Optionally generates a signed download URL for the original file (valid 5 minutes).

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
asset_h	string	required	The Asset Hash (SHA-1) of the asset to retrieve
include_download_url	boolean	optional	If true, includes a short-lived signed URL for downloading the original asset binary. Defaults to false.
include_raw_metadata	boolean	optional	If true, also returns the raw extractor output (extracted_metadata) and the node-by-node workflow_graph. These can be tens of kilobytes; the default response already carries prompts, models and workflow.workflow_json. Defaults to false.

BulkUpdateAssetTags

tool · uncategorised

Add or remove a set of tags across up to 100 assets in one call. Per-asset behaviour: 'add' merges into the existing tag set (deduped); 'remove' deletes the listed tags from each asset. Assets whose tag set does not change are skipped silently and reported as unchanged in the response. Tenant-scoped; uses the same RLS path as PATCH /api/v1/assets/bulk-tag.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
asset_hs	string[]	required	Array of asset hashes (SHA-1, 40-char hex) to apply the operation to. Min 1, max 100.
operation	string	required	'add' merges the tags into each asset's existing tag set (deduped); 'remove' deletes them. For replace-all semantics on a single asset, use UpdateAsset. enum: add · remove
tags	string[]	required	Tag strings to add or remove. Min 1 char, max 100 chars per tag. Min 1 tag, max 50 tags per call.

UpdateAsset

tool · uncategorised

Partial update of a single asset's scalar metadata: description, title, and/or tags. Tag update on this tool REPLACES the entire existing tag set; for incremental add/remove operations across multiple assets use BulkUpdateAssetTags. At least one of description, title, or tags must be provided. Tenant-scoped; uses the same RLS path as PATCH /api/v1/assets/[assetH].

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
asset_h	string	required	Asset hash (SHA-1, 40-char hex) of the asset to update.
description	string	optional	Optional new asset description (max 5000 chars). Replaces any existing description verbatim.
title	string	optional	Optional new asset title (max 500 chars).
tags	string[]	optional	Optional new tag set. REPLACES the entire existing tag set. For incremental add/remove use BulkUpdateAssetTags. Min 0 tags, max 500. Each tag 1–100 chars.

ExportCollectionAsPdf

tool · uncategorised

Export a collection as a PDF document. Returns the PDF as a base64-encoded string along with page count and processing time metadata. The web API handles layout, pagination, and optional watermark removal.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
collection_h	string	required	Collection hash (SHA-1, 40 hex chars) to export
remove_watermark	boolean	optional	Whether to remove the Numonic watermark from the exported PDF. Default: false.

AddTextCardToCollection

tool · uncategorised

Add a rich-text card (slide) to a collection. Text cards are rendered as full slides in PDF exports and can contain a heading, body text, bullet points, and an accent callout. Supports dark, gradient, and light backgrounds with optional hex accent color.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
collection_h	string	required	Collection hash (SHA-1, 40 hex chars)
heading	string	optional	Card heading text
heading_size	string	optional	Heading size: xl, lg, or md enum: xl · lg · md
body	string	optional	Body text for the card
bullets	string[]	optional	Array of bullet point strings
accent_text	string	optional	Highlighted callout text
background	string	optional	Background style. Default: dark enum: dark · gradient · light
accent_color	string	optional	Hex color code for accent elements (e.g., #3B82F6)
layout	string	optional	Layout style. Default: centered enum: centered · left-aligned
position	integer	optional	Position in collection ordering (0-based)

SearchModels

tool · uncategorised

Search the Model Library for models (LoRAs, checkpoints, UNETs, etc.) by name, type, or base architecture. Returns model entities with metadata, not assets. Use GetAssetModels to find assets by model.

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	UUID of the tenant
query	string	optional	Text search across model name, display name, and description. Supports partial matching.
model_type	string	optional	Filter by model type: lora, checkpoint, unet, clip, upscaler, controlnet, vae, embedding, ip_adapter, llm, etc.
base_model	string	optional	Filter by base architecture: sdxl, flux, sd15, sd3, sd35, pony, illustrious, noobai, hunyuan, etc.
limit	integer	optional	Maximum results to return (default: 50)
offset	integer	optional	Offset for pagination (default: 0)

GetModelDetails

tool · uncategorised

Get full details for a specific model including metadata, version history, asset usage count, and preview count.

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	UUID of the tenant
model_hash	string	required	SHA-1 hash key of the model (from models_h.model)

GetAssetModels

tool · uncategorised

Get the Model Library records for every model used to generate a specific asset (checkpoints, UNETs, LoRAs with weights, VAEs, CLIP, upscalers), with base model, type and detection source. Scoped to your tenant: an asset your tenant does not hold returns no models. For the prompt and ComfyUI workflow graph, call GetAssetDetails.

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	UUID of the tenant
asset_hash	string	required	SHA-1 hash key of the asset

LinkModelToAsset

tool · uncategorised

Manually link a model to an asset (for non-ComfyUI sources where automatic detection is not available).

Parameter	Type	Required	Description
tenant_id	string ·uuid	optional	UUID of the tenant
model_hash	string	required	SHA-1 hash key of the model
asset_hash	string	required	SHA-1 hash key of the asset
model_weight	number	optional	LoRA strength_model value (e.g., 0.75). Null for checkpoints.
detection_source	string	optional	How this link was discovered (default: manual_tag)

GetModelVersions

tool · uncategorised

Get version history for a model, including training steps, quality scores, and recommended flag.

Parameter	Type	Required	Description
model_hash	string	required	SHA-1 hash key of the model
include_deleted	boolean	optional	Include soft-deleted versions (default: false)

AddModelVersion

tool · uncategorised

Add a version to a model with training metadata, quality score, and recommended flag.

Parameter	Type	Required	Description
model_hash	string	required	SHA-1 hash key of the model
version_tag	string	required	Version label (e.g., v1.0, v2.0, epoch10)
version_notes	string	optional	Changelog or description for this version
training_steps	integer	optional	Total training steps
training_epochs	integer	optional	Total training epochs
training_config	object	optional	Training hyperparameters: {"lr": 0.0001, "optimizer": "AdamW"}
quality_score	number	optional	Quality rating (0-10)
is_recommended	boolean	optional	Mark as the recommended version for this model
trained_at	string ·date-time	optional	When this version was trained

CreateExperiment

tool · uncategorised

Create an experiment collection for structured image comparison workflows (model benchmarking, prompt A/B testing, style evaluation). Returns the experiment collection hash.

Parameter	Type	Required	Description
tenant_id	string	required	Tenant UUID
display_name	string	required	Human-readable experiment name
description	string	optional	Experiment description
hypothesis	string	optional	What this experiment tests
independent_variable	string	optional	What varies across runs (e.g., "prompt condition", "model", "LoRA")
control_variable	string	optional	What is held constant (e.g., "seed set", "resolution")
model_name	string	optional	Primary generation model (e.g., "FLUX.2 Dev")
workflow_name	string	optional	Generation workflow (e.g., "txt2img")
resolution	string	optional	Image resolution (e.g., "1024x1024")
domain_context	object	optional	Domain-specific metadata as JSON (e.g., {"grammar_version": "v3"})

RegisterRun

tool · uncategorised

Add an asset to an experiment as a run with full context (condition, seed, prompt, model, domain params).

Parameter	Type	Required	Description
collection_h	string	required	Experiment collection hash (40-char hex)
asset_h	string	required	Asset hash to register as a run (40-char hex)
condition_id	string	optional	Independent variable value for this run (e.g., "cond2b", "model_A")
variant_label	string	optional	Human-readable variant label (e.g., "Compiled prompt")
seed	integer	optional	Generation seed
prompt_text	string	optional	Exact prompt used for generation
model_name	string	optional	Model for this specific run (overrides experiment-level)
run_params	object	optional	Domain-specific run parameters as JSON
outcome_tags	string[]	optional	Outcome classification tags (e.g., ["text_literalization", "best_in_set"])
position	integer	optional	Position/order within the experiment

ScoreRun

tool · uncategorised

Score an experiment run on user-defined dimensions. Creates an evaluation annotation. Dimensions are experiment-defined (e.g., quality, adherence, coherence).

Parameter	Type	Required	Description
collection_item_h	string	required	Collection item hash for the run to score (40-char hex)
scores	object	required	Scoring dimensions as key-value pairs (e.g., {"quality": 4, "coherence": 5}). Values are 1-5.
flags	object	optional	Boolean flags (e.g., {"prompt_leakage": true, "text_literalization": false})
evaluator_type	string	optional	Who/what performed the evaluation (default: "human") enum: human · vlm_assisted · embedding_similarity
comments	string	optional	Free-form evaluation comments

CompareRuns

tool · uncategorised

Create a pairwise A/B comparison between two experiment runs, recording which item wins on each scoring dimension.

Parameter	Type	Required	Description
collection_h	string	required	Experiment collection hash (40-char hex)
item_a	string	required	First run collection item hash (40-char hex)
item_b	string	required	Second run collection item hash (40-char hex)
winners	object	required	Dimension-to-winner mapping (e.g., {"quality": "<item_a_hash>", "coherence": "<item_b_hash>"})
evaluator_type	string	optional	Who/what performed the comparison (default: "human") enum: human · vlm_assisted · embedding_similarity
notes	string	optional	Comparison notes

GetExperimentSummary

tool · uncategorised

Get aggregated experiment results: run count, condition breakdown, outcome tag distribution, and comparison count.

Parameter	Type	Required	Description
collection_h	string	required	Experiment collection hash (40-char hex)

CreatePrompt

tool · uncategorised

Create a new prompt in the Prompt Library. Returns the prompt hash and version info.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	Tenant UUID
prompt_text	string	required	The prompt text content (required)
name	string	optional	Human-readable prompt name
description	string	optional	Prompt description
target_model	string	optional	Target model (e.g., midjourney, dall-e, flux)
category	string	optional	Prompt category (e.g., landscape, portrait)
tags	string[]	optional	Prompt tags for organization

GetPrompt

tool · uncategorised

Get a prompt by hash ID with full version history and evaluations.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	Tenant UUID
prompt_h	string	required	Prompt hub hash (40-char hex)

SearchPrompts

tool · uncategorised

Search prompts with optional category, target_model filters, and pagination.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	Tenant UUID
category	string	optional	Filter by category
target_model	string	optional	Filter by target model
limit	integer	optional	Page size (default 20, max 100)
offset	integer	optional	Pagination offset (default 0)

RenderPrompt

tool · uncategorised

Render a prompt template by substituting {{variable}} placeholders with provided values.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	Tenant UUID
prompt_h	string	required	Prompt hub hash (40-char hex)
variables	object	optional	Template variables as key-value pairs (e.g., {"subject": "cityscape"})

RegisterProspect

tool · uncategorised

Register or update a cold-outreach prospect: writes prospects_h hub + tenant scoping link + prospect_profile_s, and optionally prospect_research_s (written when any research field is present, not only pain_hypothesis — #2000) and prospect_deal_s (when trial_start_date or agreed_price is present — #2000). At least one of email / linkedin_url / profile_url / handle / website required (#1997); keying precedence when more than one is present is email > linkedin_url > profile_url > handle > website. Append-only — re-calling adds new satellite rows.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
prospect	object	required	Prospect data. At least one of email / linkedin_url / profile_url / handle / website is required (validated server-side).
record_source	string	required	Origin tag, e.g. "cold-outreach-studios:studios-2026-q2".

SetProspectOwner

tool · uncategorised

Assign or reassign a prospect's owner (#1999). Append-only — writes a new prospect_owner_s row; never touches prospect_outreach_touch_s, so reassignment does not rewrite touch history. Set independently of prospect_outreach_touch_s.sent_by.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
prospect_h	string	required	Prospect hub hash (SHA-1).
owner	string	required
record_source	string	required

LogOutreachTouch

tool · uncategorised

Append a touch to the outbound log and advance the prospect outreach sequence state. channel accepts email, linkedin, discord, github, contact_form, or other (#1998). Both writes are append-only.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
prospect_h	string	required	Prospect hub hash (SHA-1).
touch	object	required
record_source	string	required

GetProspectState

tool · uncategorised

Return full prospect state: latest profile + research + state + suppression, plus all touches and replies, plus contact_link if hand-off has occurred.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
prospect_h	string	required	Prospect hub hash (SHA-1).

SearchProspects

tool · uncategorised

Filter prospects by campaign, current sequence status, minimum ICP score, vertical, account substring, or owner (#1999, exact match). Suppressed prospects (suppressed_until in the future) are excluded by default. Default limit 50, max 200. Each result row includes the current owner (null if unassigned).

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
filters	object	optional

SuppressProspect

tool · uncategorised

Suppress a prospect with a reason and optional end-date. Marks all active sequences for the prospect as suppressed. unsubscribe is ALWAYS permanent (any suppressed_until is ignored). For not_interested / wrong_person, omit suppressed_until for permanent suppression or pass a future date for a finite one. not_now requires suppressed_until (the reengage date).

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
prospect_h	string	required
reason	string	required	enum: unsubscribe · wrong_person · not_now · not_interested
suppressed_until	string,null ·date	optional	Future ISO date YYYY-MM-DD when suppression lifts. Omit (or null) for permanent suppression; required for not_now; ignored for unsubscribe (always permanent).
record_source	string	required

LogReply

tool · uncategorised

Record an inbound reply with cold-outreach-studios reply-triage classification. Advances sequence state: positive → replied, not_now/wrong_person/unsubscribe/not_interested/bounced → suppressed, ooo leaves state untouched. bounced (#1998) is a delivery failure, distinct from every refusal classification. reply.refusal_class (#2000) optionally codes a refusal T (timing) or P (problem), only valid alongside a refusal-shaped classification. On positive reply, mints prospect_to_contact_l so lead-response v0.2 picks up the now-warm contact.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
prospect_h	string	required
reply	object	required
record_source	string	required

SearchContacts

tool · uncategorised

List CRM contacts in a pipeline (community | enterprise | investor) with optional stage and free-text search. Wraps get_pipeline_contacts; returns the latest profile, pipeline stage, activity (last_contacted, next_follow_up), plus pipeline-specific satellite fields (investor: firm_name, investor_type, investment_amount, round, seis_eis_status). Default limit 50, max 200. tenant_id is auto-injected from API key auth.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
pipeline	string	required	Which CRM pipeline to list. 'investor' for fundraise contacts. enum: community · enterprise · investor
stage	string	optional	Optional pipeline-stage filter (e.g. "lead", "engaged", "diligence").
search	string	optional	Optional case-insensitive substring match against contact name and linkedin_url.
limit	integer	optional
offset	integer	optional

GetRelationshipSummary

tool · uncategorised

One answer to "who is this, who owns it, what is next, and when is it due" for a CRM contact, prospect or organization (#2637). Resolves merged duplicate contacts to the canonical record, joins hand-off prospects and the organization, reconciles owner (prospect owner > contact assigned_to > open-task assignee, all candidates returned with owner_conflict), picks the earliest-due next action across open tasks, follow-ups, active outreach touches and re-engage dates, and returns a reverse-chronological activity list (touches, replies, notes, stage changes, tasks, deal updates). Same data as the admin CRM GUI panel.

Parameter	Type	Required	Description
tenant_id	string ·uuid	required	UUID of the tenant
entity_type	string	required	Which hub entity_id belongs to. enum: contact · prospect · organization
entity_id	string	required	Hub hash (SHA-1) of the contact, prospect or organization — e.g. a contact from SearchContacts or a prospect_h from SearchProspects.
activity_limit	integer	optional	Maximum activity entries returned (default 50).

Resources(5)

Resources are browsable read-only data that MCP clients can read for context — collection structure, storage quotas, available presets. Unlike tools, resources don't modify state.

resources/list resources/read:

{"jsonrpc":"2.0","method":"resources/read","params":{"uri":"numonic://collections"},"id":3}

numonic://storage

Storage Usage

resource

Tenant storage: GB used, quota, percentage, over-limit flag. Check before uploading large batches.

numonic://export-presets

Export Presets

resource

Available privacy presets with compliance metadata (EU AI Act, CA SB 942).

numonic://pipeline-templates

Pipeline Templates

resource

System and tenant pipeline templates. Browse before building custom pipelines.

numonic://asset/{asset_h}

Asset Details

resource · template

Full metadata for a specific asset. Pass the 40-character hex asset hash. (URI template)

Prompts(5)

Guided multi-step workflows that teach agents the canonical way to accomplish common tasks. They return pre-written instructions that chain multiple tool calls together.

prompts/get:

{"jsonrpc":"2.0","method":"prompts/get","params":{"name":"search-and-curate","arguments":{"query":"tool:midjourney AND tag:approved"}},"id":5}

Ingest & Organize

ingest-and-organize prompt

Upload a new asset and file it into the correct collection.

Workflow: StoreAsset → ListCollections → AddToCollection

Argument	Required	Description
filename	required	Name of the file being uploaded
mime_type	required	MIME type (e.g., image/png)
collection_path	optional	Target collection path

Export for Client

export-for-client prompt

Export assets with privacy-aware metadata stripping for client delivery.

Workflow: SearchAssets or GetCollectionAssets → ExportAssets

Argument	Required	Description
query_or_collection	required	Search query or collection hash
preset	optional	Privacy preset (default: client)

Audit Tenant Health

audit-tenant-health prompt

Assess storage usage, search index quality, and overall tenant health.

Workflow: GetTenantStorageDetails → GetSearchAnalytics → summarize

Explore Lineage

explore-lineage prompt

Trace a Midjourney asset's complete evolution chain through variations, upscales, and remixes.

Workflow: GetMidjourneyEvolutionChain → GetCreativeSession → synthesize

Argument	Required	Description
asset_h	required	Asset hash (40-char hex)
depth	optional	Maximum chain depth (default: 10)

Error Codes

The MCP server uses standard JSON-RPC 2.0 error codes.

Code	Name	Meaning
-32700	Parse Error	Malformed JSON in request body
-32600	Invalid Request	Missing jsonrpc or method field
-32601	Method Not Found	Unknown method name
-32602	Invalid Params	Missing or invalid parameters
-32603	Internal Error	Server-side exception

HTTP Status Codes

Parse errors and invalid requests return HTTP 400. All other errors (including tool failures) return HTTP 200 with the error in the JSON-RPC response body — per the JSON-RPC specification.

Tool-Level Errors

Tool errors are returned in the result's content array with isError: true:

{"content":[{"type":"text","text":"Error: …"}],"isError":true}

Protocol Details

Protocol	Model Context Protocol (JSON-RPC 2.0)
Transport	Streamable HTTP (POST)
Supported versions	2025-03-26 (primary), 2024-11-05 (backwards compatible)
Server name	Numonic-MCP-Server
Server version	1.0.0
Tools	66
Resources	5 (4 static + 1 URI template)
Prompts	5 guided workflows

MCP Server Reference: Tools, Resources & Client Setup