alchemy-automation

Overview of alchemy-automation skill

What alchemy-automation does

alchemy-automation is a Claude skill for running Alchemy-related workflows through Composio’s Rube MCP server. Instead of assuming fixed API shapes, the skill tells the agent to search Rube tools first, inspect the current Alchemy toolkit schemas, confirm the account connection, and then execute the appropriate tool calls.

This makes the alchemy-automation skill useful when you want an AI agent to help with Alchemy operations without manually browsing Composio tool documentation every time.

Best-fit users and jobs

Use this skill if you already work with Alchemy and want Claude to operate through an MCP tool layer rather than only giving written instructions. It is a good fit for developers, DevOps users, blockchain teams, and automation builders who need an agent to discover available Alchemy actions, prepare valid inputs, and follow a repeatable execution pattern.

The real job is not “explain Alchemy.” It is: connect Rube MCP, verify the Alchemy toolkit connection, discover the exact tool schema for the current task, and run the workflow with less schema guesswork.

Key differentiator: schema discovery first

The main value of alchemy-automation is its insistence on RUBE_SEARCH_TOOLS before execution. That matters because MCP tool schemas can change, and generic prompts often fail by inventing parameters or calling outdated tool names.

The skill’s workflow is intentionally conservative: search tools, check connection, review schemas and pitfalls, then execute. That is the difference between a prompt that describes an automation and a skill that can drive one through Rube MCP.

How to Use alchemy-automation skill

alchemy-automation install and setup context

Install the skill from the Composio skills repository, for example:

npx skills add ComposioHQ/awesome-claude-skills --skill alchemy-automation

Then configure Rube MCP in your AI client by adding:

https://rube.app/mcp

The upstream skill expects RUBE_SEARCH_TOOLS to be available and requires an active Alchemy connection through RUBE_MANAGE_CONNECTIONS with toolkit alchemy. If the connection is not active, the agent should follow the returned authorization link before attempting any Alchemy operation.

What to read before using it

Start with composio-skills/alchemy-automation/SKILL.md. This repository entry is compact and has no extra README.md, scripts, references, or rule folders, so the skill file is the source of truth.

Pay particular attention to these sections:

• Prerequisites for required MCP and connection state
• Setup for adding Rube MCP and activating Alchemy
• Tool Discovery for the required RUBE_SEARCH_TOOLS call
• Core Workflow Pattern for the order of operations

Because there are no helper scripts, your results depend heavily on giving the agent a precise task and letting it discover current Rube schemas.

How to prompt the skill well

A weak prompt is:

“Use Alchemy to do this for me.”

A stronger alchemy-automation usage prompt includes the operation, target network or account context, required output, and safety boundaries:

“Use the alchemy-automation skill through Rube MCP. First call RUBE_SEARCH_TOOLS for the current Alchemy schema. Then check my alchemy connection status. I need to perform [specific Alchemy task] for [network/project/account context]. Do not execute destructive or billing-affecting changes without summarizing the planned tool call and asking for confirmation.”

This gives the agent enough structure to invoke the skill correctly while still relying on Rube for the current tool list and parameters.

Suggested workflow for reliable execution

Use this sequence for most tasks:

1. Ask the agent to search tools with a use case that matches your actual task, not a broad phrase like “Alchemy operations.”
2. Have it confirm alchemy connection status with RUBE_MANAGE_CONNECTIONS.
3. Ask it to summarize the discovered tool slug, required fields, optional fields, and known pitfalls.
4. Provide missing values only after the schema is known.
5. For high-impact actions, require a dry-run-style summary before execution.

This workflow is slower than a one-shot prompt, but it reduces failed calls caused by stale tool assumptions.

alchemy-automation skill FAQ

Is alchemy-automation only for Composio users?

It is specifically built around Composio’s Rube MCP server and the Composio Alchemy toolkit. If your AI client cannot connect to MCP tools or does not expose RUBE_SEARCH_TOOLS and RUBE_MANAGE_CONNECTIONS, this skill will not be very useful beyond showing a workflow pattern.

How is it better than an ordinary Alchemy prompt?

An ordinary prompt can explain Alchemy concepts or draft code, but it cannot reliably know the current Rube MCP tool schemas. The alchemy-automation skill instructs the agent to discover available tools at runtime, check authentication state, and use the returned schema rather than hallucinating parameters.

That makes it better for live tool execution, not necessarily for tutorials or general blockchain education.

Is this skill beginner-friendly?

It is beginner-friendly only if Rube MCP and the Alchemy connection are already configured. New users may still need help understanding MCP setup, Alchemy authorization, and which Alchemy operation they actually want.

For best results, beginners should ask the agent to explain the discovered tool schema before running it.

When should I not use this skill?

Do not use alchemy-automation when you only need static documentation, SDK code generation, or a conceptual explanation of Alchemy. Also avoid it for tasks where you cannot authorize the Alchemy toolkit, cannot inspect tool calls before execution, or need a custom workflow unsupported by the current Composio toolkit.

How to Improve alchemy-automation skill

Make alchemy-automation prompts more specific

The biggest improvement is better task framing. Replace broad goals with operational details:

• target Alchemy product or workflow
• network, app, project, or account context
• required output format
• whether execution is allowed or only planning
• confirmation rules for risky actions

For example:

“Search Rube tools for managing Alchemy app settings. Identify the correct tool, list required inputs, and wait for my confirmation before making changes.”

This produces better results than asking the agent to “handle Alchemy automation.”

Prevent common failure modes

Common failures include skipping tool discovery, assuming stale schemas, trying to execute before the alchemy connection is active, and asking for a tool call without enough required fields.

To reduce these failures, explicitly require:

• RUBE_SEARCH_TOOLS before any execution
• RUBE_MANAGE_CONNECTIONS status check
• a schema summary before collecting inputs
• confirmation before irreversible, security-sensitive, or billing-related operations

These constraints align with the skill’s own design rather than fighting it.

Iterate after the first tool search

The first search may return several possible tools or an execution plan that is close but not exact. After the first result, refine the prompt with the discovered terminology.

Instead of continuing with your original wording, say:

“Use the discovered tool slug that supports [specific action]. Fill only fields confirmed by the schema. Ask me for any missing required values.”

This keeps the workflow grounded in the actual Rube response.

Improve the upstream skill for teams

Teams adopting the alchemy-automation skill can make it stronger by adding local examples for their most common Alchemy tasks, confirmation policies, and approved networks or projects. Since the repository version contains only SKILL.md, team-specific usage notes can reduce repeated setup questions and prevent accidental execution in the wrong environment.

Good additions include sample prompts, safe-operation checklists, and internal rules for when the agent must stop and request human approval.