Core Concepts
This page explains the key concepts you need to understand to use Wire effectively.
Containers
Section titled “Containers”A Container is the fundamental unit of Wire. Each container has its own MCP server, content, and access controls.
Container Properties
Section titled “Container Properties”| Property | Description |
|---|---|
| Name | Human-readable name for the container |
| Visibility | private (requires auth) or public (open access) |
| Paused | When paused, MCP requests are blocked |
| Sources | Files uploaded to the container |
| Tools | Auto-generated MCP tools for accessing context |
Container Lifecycle
Section titled “Container Lifecycle”- Created - Container is provisioned
- Initializing - Processing source changes
- Ready - Container accepts MCP queries
- Paused - MCP requests blocked (can be resumed)
- Deleted - Container and contents removed
Sources
Section titled “Sources”Sources are the content you add to containers. Currently, files are the only supported source type.
Supported File Formats
Section titled “Supported File Formats”- Documents: PDF, Word (.docx), Text, Markdown
- Data: JSON, CSV
See Supported File Types for the complete list.
How Files Become Context
Section titled “How Files Become Context”When you upload files, Wire processes them into queryable context. Depending on the content, this happens in different ways:
Unstructured content (documents, books, articles) gets chunked for semantic search. The content remains searchable as text.
Structured data (CSV, JSON) gets transformed to extract entities and records. Wire identifies the key entities and creates queryable records from them.
Mixed content can be both - entities are extracted for structured queries while the content remains available for semantic search.
Examples
Section titled “Examples”| Container | Content | Result |
|---|---|---|
| UN Demographics | Multiple CSV files | Entities like countries and years extracted, records connected across files |
| Chocolate Recipes | Recipe book PDF | Recipes extracted as entities for structured queries + full semantic search |
| A Christmas Carol | Novel by Dickens | Semantic searchable content |
Processing Pipeline
Section titled “Processing Pipeline”When you upload a file, Wire:
- Validates - Checks format and size limits
- Stores - Saves the file securely
Then, depending on the content:
- Chunks - Splits unstructured content for semantic search
- Transforms - Extracts entities and records from structured data
- Both - Some files get chunked and transformed
MCP Tools
Section titled “MCP Tools”Each container automatically generates MCP tools for AI agents:
Standard Tools
Section titled “Standard Tools”| Tool | Description |
|---|---|
list | List entities or records in the container |
get | Retrieve a specific record |
filter | Filter records by criteria |
search | Search across records |
semantic_search | AI-powered search (may return file content) |
Tool Costs
Section titled “Tool Costs”Wire uses a credit system for tool calls. See the pricing page for current rates.
Organizations
Section titled “Organizations”Organizations contain users, containers, and billing information.
| Role | Capabilities |
|---|---|
| Owner | Full control, billing, delete org |
| Admin | Manage members, create containers |
| Member | Use containers, read-only settings |
See Roles & Permissions for details.
Authentication
Section titled “Authentication”Containers can be public or private (default).
- Public containers are accessible without authentication
- Private containers require OAuth or API keys
For private containers, Wire uses OAuth 2.1:
- AI tool initiates OAuth flow
- User authorizes in browser
- Token issued with container scopes
- AI tool makes authenticated requests
See MCP Authentication for details.