borneo-automation

Overview of borneo-automation skill

What borneo-automation is for

borneo-automation is a Claude skill for running Borneo-related operations through Composio’s Rube MCP server. Its main value is not a fixed list of hard-coded actions; it teaches the agent to discover the current Borneo toolkit tools first, verify the user’s connection, then execute with the latest schemas returned by Rube.

Use this skill when you want an AI agent to operate Borneo through MCP instead of manually checking Composio toolkit docs, guessing tool names, or writing one-off prompts that may use outdated parameters.

Best-fit users and workflows

The borneo-automation skill is best for users who already have, or can create, a Borneo connection in Composio/Rube and want Claude to help with repeatable workflow automation. It fits operations where the exact tool schema may change, because the skill explicitly requires RUBE_SEARCH_TOOLS before execution.

It is a good match for:

• Teams using Claude with MCP-enabled clients.
• Users automating Borneo tasks through Composio’s Borneo toolkit.
• Workflows where connection status, auth state, and tool schema accuracy matter.
• Agents that need a reliable discovery-check-execute pattern.

What makes this skill different

The important differentiator is its schema-first workflow. Instead of assuming a tool name or stale parameter structure, borneo-automation instructs the agent to call RUBE_SEARCH_TOOLS, inspect available Borneo tools, then use RUBE_MANAGE_CONNECTIONS to confirm the borneo toolkit connection is active.

That makes the skill useful for real workflow automation where MCP tools evolve over time. The tradeoff is that the skill is intentionally thin: it does not include scripts, local helpers, sample business workflows, or embedded reference files. Its quality depends on Rube MCP availability and the live Composio Borneo toolkit response.

How to Use borneo-automation skill

borneo-automation install and MCP setup

Install the skill from the source repository with:

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

Then make sure your Claude-compatible client has Rube MCP configured. The skill’s source points to:

https://rube.app/mcp

After MCP is available, confirm that RUBE_SEARCH_TOOLS appears as a callable tool. Then use RUBE_MANAGE_CONNECTIONS with toolkit borneo to check whether the connection is ACTIVE. If Rube returns an auth link, complete authorization before asking the agent to run Borneo actions.

Inputs the skill needs from you

For good borneo-automation usage, do not ask only “do the Borneo task.” Give the agent enough operational context to choose and validate the right tool after discovery.

A weak prompt:

Use Borneo to update the record.

A stronger prompt:

Use borneo-automation for Workflow Automation. First discover current Borneo tools with RUBE_SEARCH_TOOLS, then check my borneo connection. I need to update the Borneo item matching [identifier] so that [field] becomes [new value]. If multiple tools or records match, show the options and ask before executing.

Useful inputs include:

• The specific Borneo task or business outcome.
• Known IDs, names, filters, dates, or record fields.
• Whether the agent may execute changes or should only prepare a plan.
• How to handle ambiguity, duplicates, missing records, or failed auth.

Recommended workflow inside Claude

A reliable borneo-automation guide should follow this sequence:

1. Ask Claude to use the installed borneo-automation skill.
2. Have it call RUBE_SEARCH_TOOLS with your specific use case, not a generic query.
3. Have it inspect returned tool slugs, schemas, execution plans, and pitfalls.
4. Confirm borneo connection status with RUBE_MANAGE_CONNECTIONS.
5. Execute only after required fields are known and the connection is active.
6. Ask for a short execution summary, including tool used, inputs, result, and any unresolved items.

If your task changes mid-conversation, ask Claude to search again. The upstream skill is explicit that current schemas should be discovered first.

Repository files to read first

This skill currently consists of SKILL.md only. Read that file before install if you want to verify scope, prerequisites, and the MCP pattern. There are no bundled scripts, README.md, references, rules, or sample workflow folders in the skill directory, so the live Rube search results are the real source of tool-level detail.

For external context, the source points to the Composio Borneo toolkit documentation at composio.dev/toolkits/borneo.

borneo-automation skill FAQ

Is borneo-automation useful without Rube MCP?

No. The skill requires Rube MCP, specifically access to tools such as RUBE_SEARCH_TOOLS and RUBE_MANAGE_CONNECTIONS. Without MCP tool access, it becomes only a prompt pattern and cannot automate Borneo operations.

How is this better than an ordinary prompt?

An ordinary prompt may guess Borneo tool names, parameters, or auth state. The borneo-automation skill forces a safer sequence: discover tools, verify connection, then execute using the current schema. That reduces failures caused by stale docs, renamed tools, or missing required fields.

Is this suitable for beginners?

It can be beginner-friendly if your MCP client is already configured. The setup concept is simple, but users must understand that the agent needs live tool access and an active Borneo connection. If you are new to MCP, expect the first session to focus on connection setup rather than task execution.

When should I not use this skill?

Do not use borneo-automation when you need offline processing, local-only scripts, or a fully documented Borneo workflow template included in the repository. Also avoid it if your organization does not allow agents to call external MCP tools or manage connected app authorization.

How to Improve borneo-automation skill

Improve prompts for borneo-automation

The fastest way to improve results is to make the desired operation testable. Include the target object, selection criteria, allowed action, and confirmation rules.

Better prompt pattern:

Use borneo-automation. Search current Borneo tools for [exact task]. Check my borneo connection. If the matching tool requires fields I have not provided, ask for them before execution. Do not modify data until I confirm the planned tool call.

This helps the agent avoid premature execution and makes schema discovery useful instead of ceremonial.

Common failure modes to prevent

The most common failure is skipping tool discovery and using assumed schemas. Another is trying to execute while the Borneo connection is inactive. A third is giving vague record references that cause the agent to select the wrong object.

Prevent these by requiring:

• RUBE_SEARCH_TOOLS before every new task type.
• RUBE_MANAGE_CONNECTIONS before write actions.
• A confirmation step for destructive or irreversible operations.
• Explicit handling for “no match” and “multiple matches.”

Iterate after the first output

After the first execution or plan, ask Claude to report what it actually learned from Rube: tool slug, required inputs, optional inputs, returned warnings, and execution result. If the outcome is incomplete, refine the next prompt around the missing field or failed assumption rather than restarting broadly.

Example follow-up:

Re-run discovery only if needed. Use the same session if possible. The previous attempt failed because [field] was missing. Ask me only for fields required by the selected Borneo tool, then continue.

Practical additions for teams

Teams can improve this skill by adding local examples, approved workflow prompts, risk rules for write actions, and organization-specific naming conventions. A small references/ file with common Borneo tasks and required business context would make borneo-automation more repeatable without weakening its schema-first design.