Hydra ACI
The Hydra ACI server acts as a web capability compiler that discovers, compiles, and exposes any website's machine-usable surfaces as typed, callable tools for AI agents — no site-side API or MCP server required.
navigate— Navigate the stealth browser to any URL and discover session-specific actions available on that page.hydra_execute— One-shot autonomous task execution: provide a URL and a goal, and Hydra autonomously completes the entire task (filling forms, calling APIs, navigating) with no human in the loop except for payment/MFA challenges.hydra_mount— Mount a website as a persistent workspace, triggering full discovery (OpenAPI, GraphQL, XHR, DOM, MCP apps, embedded data) and returning a canonical typed tool catalog; optionally accepts credentials and a goal for route planning.hydra_probe— Mount a site for browser-as-probe discovery, returning direct machine-usable capabilities separately from browser-action fallbacks — without executing any discovered capability.hydra_refresh— Re-scout an already-mounted site and refresh its capability catalog to pick up any changes or drift.hydra_call_tool— Call any capability in the mounted catalog by its canonical ID or name, passing structured arguments across OpenAPI endpoints, XHR calls, forms, and browser actions through one unified interface.hydra_inspect_session— Inspect the current mounted session state and active route plan.hydra_memory— Manage the H8 Knowledge Archive for a domain: retrieve stats, export stored workflows, or forget all memory for a domain to force fresh discovery on next mount.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Hydra ACI mount https://github.com and list features"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
safe source snapshot
Available Tools
8 toolshydra_call_toolB
Call a mounted Hydra capability by canonical id, namespaced capability name, or original tool name.
Inputs: tool_name (required), arguments.
hydra_executeB
One-shot autonomous execution: provide a target URL and goal, and Hydra will complete the task automatically.
Inputs: goal (required), target (required).
hydra_inspect_sessionA
Inspect the current mounted Hydra session and route plan.
Inputs: no named parameters.
hydra_memoryC
Inspect or manage H8 web memory for a domain.
Inputs: action (required), domain.
hydra_mountA
Mount a website as a persistent Hydra web capability workspace and return its canonical capability catalog.
Inputs: target (required), goal, credentials.
hydra_probeC
Mount a website for browser-as-probe discovery, then return direct machine-usable capabilities separately from browser-action fallbacks without executing a generated capability.
Inputs: target (required), goal, credentials.
hydra_refreshA
Re-scout the mounted site and refresh its capability catalog.
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 8 tools
The tools have distinct purposes, but 'hydra_mount' and 'hydra_probe' both involve mounting and could cause confusion despite different intents. Overall, agents can distinguish them with careful reading.
Most tools use the 'hydra_' prefix, but 'navigate' breaks the pattern. The naming styles vary: some are single verbs, some verb_noun, and 'hydra_memory' is a noun. This inconsistency reduces predictability.
With 8 tools, the count fits well within the recommended 3-15 range. Each tool appears justified for the server's complex web automation domain.
The tool set covers mounting, probing, refreshing, inspecting, memory management, calling capabilities, executing tasks, and navigation. A notable gap is the lack of an explicit unmount or close tool, but the core workflow is covered.
Related MCP Connectors
Deterministic Zero-LLM Web Compiler, Scraper, Crawler, and Site Mapper for AI Agents.
AI-powered web automation. Navigate websites using AI agents for one page or a thousand
AI-powered web automation. Navigate websites using AI agents for one page or a thousand
Web scraping, browser automation, screenshots and AI data extraction for AI agents via MCP.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables real browser automation as tools in Cursor, Claude Desktop, Windsurf, and any MCP-compatible client, allowing AI agents to interact with web pages through natural language.7 npmMIT
- FlicenseNot gradedqualityDmaintenanceProvides 42+ MCP tools for browser automation, web scraping, and search, enabling AI agents like Claude and Cursor to browse, extract data, and run research agents on the live web.8-
- AlicenseNot gradedqualityBmaintenanceMCP server that turns any website into an API by capturing or importing API endpoints, enabling AI agents to interact with web services without a browser, with 20-100x token cost reduction versus browser automation.132 npm132Apache 2.0
- FlicenseNot gradedqualityCmaintenanceTurns any website into a rich set of MCP tools for scraping, crawling, structured data extraction, and automatic API mounting via OpenAPI specs.-