These files allow AI agents, LLMs, and developer tools to consume the Contit API programmatically.
MCP Server
To connect Claude, ChatGPT, Gemini, Claude Code or other MCP clients to your workspaces, use the Contit MCP servers: sign in with your Contit account, or use a Personal Access Token.
OpenAPI Specification
The full API specification in machine-readable format:
| File | Format | Use case |
|---|---|---|
| openapi.yaml | YAML | Import in Postman, Insomnia, Swagger UI |
| openapi.json | JSON | Programmatic consumption, code generators |
LLMs.txt
| File | Description |
|---|---|
| llms.txt | Structured index of the documentation for AI agents (llmstxt.org standard) |
AI agents can fetch this file to understand what the Contit API offers and find the right documentation page.
Tool Schemas
Pre-built tool definitions ready to use with AI platforms:
| File | Platform | Use case |
|---|---|---|
| claude.json | Claude (Anthropic) | Use Contit API endpoints as Claude tools |
| openai.json | OpenAI (GPT) | Use Contit API endpoints as GPT function calls |
Example: Using with Claude
python
import json
import anthropic
# Load Contit tool definitions
with open("claude.json") as f:
contit_tools = json.load(f)["tools"]
client = anthropic.Anthropic()
response = client.messages.create(
model="claude-sonnet-4-20250514",
tools=contit_tools,
messages=[{"role": "user", "content": "Search all published articles"}]
)
Example: Using with OpenAI
python
import json
import openai
with open("openai.json") as f:
contit_tools = json.load(f)["tools"]
client = openai.OpenAI()
response = client.chat.completions.create(
model="gpt-4o",
tools=contit_tools,
messages=[{"role": "user", "content": "Search all published articles"}]
)