Parallel Data Enrichment: Install, Source and Security | FunnelSlayer

Parallel Data Enrichment

Published by harshanandak in forge

Review recommended11 installs

What this skill does

Structured company and entity data enrichment using Parallel AI Task API with core/base processors. Returns typed JSON output. No binary install — requires PARALLEL_API_KEY in .env.local.

Add Parallel Data Enrichment to your agent

Review the source and files first. When you are ready, copy the prompt instruction or use the CLI command supported by your environment.

Install with a prompt

Paste this into a compatible coding agent:

add this skill "parallel-data-enrichment" from https://github.com/harshanandak/forge

Install with the CLI

Run this command in a controlled environment after reviewing the repository:

npx skills add https://github.com/harshanandak/forge --skill parallel-data-enrichment

Skill instructions

Parallel Data Enrichment

Enrich company or entity data into structured JSON using the Task API. Use core (1-5 min, $0.025) or base (15-100s, $0.01) for structured output.

CLI alternative (recommended): Install parallel-cli for official skill: npx skills add parallel-web/parallel-agent-skills --skill parallel-data-enrichment

Setup

API_KEY=$(grep "^PARALLEL_API_KEY=" .env.local | cut -d= -f2)

Create Enrichment Task

curl -s -X POST "https://api.parallel.ai/v1beta/tasks/runs" \
  -H "x-api-key: $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": "OpenAI",
    "processor": "core",
    "output_schema": {
      "type": "object",
      "properties": {
        "name": {"type": "string"},
        "founded_year": {"type": "integer"},
        "headquarters": {"type": "string"},
        "employee_count": {"type": "integer"},
        "key_products": {"type": "array", "items": {"type": "string"}}
      }
    }
  }'

Response: {"id": "task_abc123", "status": "queued"}

Check Result

curl -s "https://api.parallel.ai/v1beta/tasks/runs/task_abc123" \
  -H "x-api-key: $API_KEY"
{
  "id": "task_abc123",
  "status": "completed",
  "result": {
    "content": {
      "name": "OpenAI",
      "founded_year": 2015,
      "headquarters": "San Francisco, CA",
      "employee_count": 770,
      "key_products": ["ChatGPT", "GPT-4", "DALL-E", "Whisper"]
    },
    "basis": {
      "citations": [{"url": "...", "excerpt": "..."}]
    }
  }
}

Processors

ProcessorSpeedCostUse For
base15-100s$0.01/taskQuick lookups, simple data
core1-5 min$0.025/taskEnrichment, verification, structured data

Tips for output_schema

  • Keep schemas simple — fewer fields = more reliable
  • Use "type": "string" broadly; avoid strict enums
  • Omit optional fields from the schema

When to Use

  • Company or person data enrichment
  • Structured data extraction with typed schemas
  • Lead qualification, CRM enrichment, research

For narrative reports, use parallel-deep-research. For web search, use parallel-web-search.

Files included

  • SKILL.md

More skills