What is Archal?

Archal provides testing sandboxes for AI agents and integration engineers. Each sandbox contains one or more stateful API environments that simulate services like GitHub, Slack, Stripe, Linear, and Supabase. Environments expose familiar REST and MCP endpoints, retain mutations between calls, and reset to a declared baseline, so tests and rollouts are repeatable without shared test accounts or rate limits.

How to use Archal?

1. Sign up and get $20 in free usage credits (no card required). 2. Create a sandbox and choose the API environments it needs, with a sample, custom, or blank starting state. 3. Receive isolated REST and MCP endpoints for each environment. 4. Call the APIs using familiar paths and payloads, then inspect requests, responses, mutations, timing, and usage. 5. Reset the sandbox to its declared baseline and repeat. Coding agents can also be set up via the Archal CLI skill for Claude Code, Codex, Cursor, or Devin.

Core features of Archal

Stateful API environments that persist changes between calls and reset to a declared baseline
Isolated REST and MCP endpoints per environment
One API key to create and manage every sandbox and environment
Support for GitHub, Slack, Stripe, Linear, Supabase, Datadog, Jira, HubSpot, and more
Request/response inspection with mutations, timing, and usage tracking
Fast provisioning and reset (sandbox in ~690 ms, reset in ~245 ms)
Coding agent setup via CLI skill for Claude Code, Codex, Cursor, and Devin
Use cases for harness testing, CI integration testing, and agent training/RL