Skip to content
← Tools / Skills
ToolPica/10 trustMetrics-checked

Pica

One API connecting AI agents to 100+ third-party services

Visit official site →

About

Pica is a unified API platform that connects AI agents to third-party services such as CRMs, email, calendars, and databases. You connect accounts once in the Pica dashboard, and the agent accesses them through tool calling; the vendor claims more than 200 supported platforms. It exposes its capabilities through an MCP server distributed as @picahq/mcp, plus SDKs and a CLI for managing connections.

Vendor-published integration guides cover the OpenAI Agents SDK, Vercel AI SDK, Mastra, and CrewAI. Authentication runs through a PICA_SECRET key, and the company positions Pica as the managed auth and action layer so agent builders do not hand-roll per-service OAuth.

What it can do

  • Unified API across 200+ third-party platforms (vendor-claimed)
  • MCP server for agent tool calling
  • Managed OAuth and credential storage
  • Prebuilt skills for OpenAI Agents SDK, Vercel AI SDK, Mastra, CrewAI
  • CLI for account connections
  • Per-connection scoping

Setup

  1. Sign up at picaos.com and get a secret key from the dashboard
  2. Connect your accounts (Gmail, CRM, etc.) under Connections
  3. Add the MCP server to your agent config: npx @picahq/mcp with PICA_SECRET set
  4. Let the agent discover and call the connected tools

Prerequisites

A Pica account with API key, and an agent framework with MCP client support.

Use cases

agent-tool-useapi-integrationsmcp-serversthird-party-connectors

Works with

Trust score — 6/10

/10 trustMetrics-checked

Scored 6/10 (TIER metrics, metrics-v1). Evidence: live product with public docs at docs.picaos.com; vendor-published integration skills for major agent SDKs; claims 200+ supported platforms. Strengths: MCP-native design, broad integration catalog. Docked for: (1) closed source with no public adoption metrics; (2) integration coverage and reliability are vendor-claimed and not independently verified. License: proprietary.

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
6.0/10
Maintenance · recency, releases, issue load
7.0/10
Openness · license + docs
4.0/10
Manual review · sanity read + red-flag check
6.0/10
  • · Live product; public docs at docs.picaos.com
  • · Vendor claims 200+ supported platforms (unverified)
  • · Vendor-published integration skills for OpenAI Agents SDK, Vercel AI SDK, Mastra, CrewAI
  • · No public repo (picahq/pica returns 404); no public adoption metrics

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.