CatchAll API SKILL

The CatchAll API SKILL is a structured prompt file that teaches an AI agent how to use the CatchAll API. It contains the complete API workflow — authentication, endpoints, query construction, polling, and result formatting — so the agent can make API calls directly without additional tooling. For Claude-specific setup with MCP, see the Claude integration page.

How it works

The API SKILL file is a Markdown document that an agent reads as instructions. When the agent has code execution capabilities (shell commands, HTTP requests, or SDK calls), it uses the API SKILL to construct and send API requests on its own. No connector, plugin, or middleware is required.

Diagram showing an AI agent with code execution reading SKILL.md and calling the CatchAll API server directly over HTTPS.

The API SKILL is platform-agnostic. It works with any agent runtime that supports Markdown-based instructions and can execute HTTP requests, including Claude (claude.ai, Claude Desktop, Claude Code), OpenAI Assistants, LangChain agents, and custom agent frameworks.

Before you start

CatchAll API key from platform.newscatcherapi.com
An AI agent with code execution capabilities (for example, Claude with computer use, an OpenAI Assistant with Code Interpreter, or a custom agent with HTTP access)

API SKILL contents

The API SKILL archive contains:

CatchAll-SKILL

SKILL.md

references

VALIDATORS.md

ENRICHMENT-TYPES.md

MONITOR-SCHEDULING.md

openapi-spec.json

assets

example-submit.json

example_pull_response.json

File	Purpose
`SKILL.md`	Main prompt: authentication, job workflow (submit → poll → pull), monitor workflow, result presentation rules, and edge case handling
`references/VALIDATORS.md`	Guidance on writing effective validators, including cost control, the validators-vs-date-range distinction, and common validator patterns
`references/ENRICHMENT-TYPES.md`	Available enrichment types (`text`, `number`, `date`, `option`, `url`, `company`) with real-world examples
`references/MONITOR-SCHEDULING.md`	Natural language schedule formats and webhook configuration
`references/openapi-spec.json`	Full OpenAPI 3.0 specification for the CatchAll API
`assets/example-submit.json`	Minimal example submit request body (query and limit)
`assets/example_pull_response.json`	Example pull response with clustered results

Set up

How you load the API SKILL depends on your agent platform. The core requirement is the same everywhere: the agent must be able to read SKILL.md and its reference files, and have access to your API key.

Claude (claude.ai)
Python agent (any LLM)

Claude supports SKILL files natively through the Skills feature.

Download the API SKILL archive

Download CatchAll-SKILL.zip from the integrations repository.

Upload as a skill

Go to claude.ai/customize/skills. Click +, select Upload a skill, and choose CatchAll-SKILL.zip.Claude installs the API SKILL and applies it to all your conversations and projects.

Set your API key

When you run your first query, the API SKILL instructs Claude to make API calls with an X-API-Key header. Provide your API key when prompted, or set it as an environment variable in your project.

When the API SKILL is active, Claude uses its code execution environment to call the CatchAll API directly over HTTPS. No MCP connector is needed for this path.

Load SKILL.md as the system prompt in your agent framework.

from pathlib import Path

skill_prompt = Path("CatchAll-SKILL/SKILL.md").read_text()

# Use as system prompt with any LLM SDK
messages = [
    {"role": "system", "content": skill_prompt},
    {"role": "user", "content": "Find fintech funding rounds in the last 7 days, limit 10"}
]

The agent needs a tool or function that can execute HTTP requests. The API SKILL contains the base URL https://catchall.newscatcherapi.com, required header X-API-Key, and request formats for each endpoint.

Example prompts

Describe your research task in natural language. The agent submits the job, tracks progress, and returns structured results.

Find Series B fintech funding rounds announced in the last 7 days, limit 10.

Search for AI chip export restrictions in the last 30 days and summarize the key regulatory actions.

Find pharma acquisitions over $1B this month, then create a table sorted by deal value.

Include a date range and a record limit in your prompt to control scope and cost. For example: “in the last 14 days, limit 20”.

Write effective queries

Get better results from CatchAll jobs

API reference

Full endpoint documentation

Claude integration

MCP server setup and Python agent examples for Claude

SKILL source

SKILL.md and reference files on GitHub

Get started

Guides and concepts

How to

API Reference

Libraries

Integrations

How it works

Before you start

API SKILL contents

Set up

Example prompts

See also

Write effective queries

API reference

Claude integration

SKILL source

Get started

Guides and concepts

How to

API Reference

Libraries

Integrations

​How it works

​Before you start

​API SKILL contents

​Set up

​Example prompts

​See also

Write effective queries

API reference

Claude integration

SKILL source

How it works

Before you start

API SKILL contents

Set up

Example prompts

See also