landbot-automation
by ComposioHQlandbot-automation helps agents automate Landbot operations through Composio’s Landbot toolkit via Rube MCP, with connection checks and schema-first tool discovery before execution.
This skill scores 66/100, which makes it acceptable but limited for directory listing. Directory users get a usable trigger and connection workflow for Landbot automation through Rube MCP, but should view it as a thin operational wrapper rather than a complete Landbot playbook with concrete automation recipes.
- Clear prerequisite model: requires Rube MCP, an active Landbot connection, and use of RUBE_SEARCH_TOOLS before execution.
- Provides a concrete setup sequence for adding the Rube MCP endpoint, managing the Landbot connection, and confirming ACTIVE status.
- The tool-discovery-first pattern should help agents avoid stale schemas when automating Landbot via Composio.
- No support files, examples, or install metadata beyond the SKILL.md; users must already know how to configure MCP servers in their client.
- The workflow guidance is mostly generic Rube discovery/connection handling rather than Landbot-specific task recipes, so execution still depends heavily on live tool search results.
Overview of landbot-automation skill
What landbot-automation does
landbot-automation is a Claude skill for automating Landbot tasks through Composio’s Landbot toolkit using Rube MCP. Its main value is not a fixed Landbot script; it gives the agent a safe operating pattern: connect Rube MCP, verify the Landbot account connection, search for current tool schemas, then execute the right Landbot action with the discovered parameters.
This matters because Landbot automation often fails when an agent guesses stale API fields. The skill explicitly tells the agent to call RUBE_SEARCH_TOOLS first, so the workflow adapts to the current Composio tool definitions.
Best-fit users and workflows
The landbot-automation skill is best for users who want an AI agent to help with operational Landbot work, such as inspecting available Landbot actions, preparing bot-related changes, managing workspace objects, or chaining Landbot steps into a broader workflow automation process.
It fits teams already using Claude-compatible skills and MCP tools, especially if they want Landbot actions to be executed through Composio rather than custom API code. It is less useful if you only need copywriting for a chatbot script, because the skill is about tool-backed Landbot operations, not conversational design alone.
Key differentiator: schema-first execution
The strongest differentiator is the “search tools first” rule. Instead of assuming tool names or input shapes, the skill directs the agent to use:
RUBE_SEARCH_TOOLS
with a specific Landbot use case. That search returns available tool slugs, input schemas, suggested execution plans, and pitfalls. For workflow automation, this lowers the risk of malformed tool calls, missing required fields, and broken assumptions about what the Landbot toolkit currently supports.
How to Use landbot-automation skill
landbot-automation install and setup context
To install from the source repository in a Claude skills environment, use the repository path for the skill:
npx skills add ComposioHQ/awesome-claude-skills --skill landbot-automation
Then configure Rube MCP in your client by adding:
https://rube.app/mcp
The skill requires Rube MCP access and a Landbot connection through Composio. Before asking for real automation, confirm that RUBE_SEARCH_TOOLS is available. Then use RUBE_MANAGE_CONNECTIONS with toolkit landbot. If the connection is not ACTIVE, complete the returned authorization flow and re-check status before proceeding.
Inputs the skill needs to work well
A good landbot-automation usage prompt should include the specific Landbot outcome, workspace context, constraints, and how cautious the agent should be. Weak prompt:
“Update my Landbot flow.”
Stronger prompt:
“Use landbot-automation for Workflow Automation. First discover current Landbot tools with Rube. Check whether my Landbot connection is active. I want to list available bots, identify the bot named Lead Qualification, and prepare the safest available update path for changing the welcome message. Do not modify anything until you show the discovered tool schema and proposed action.”
This works better because it separates discovery, connection validation, planning, and execution permission.
Recommended workflow for reliable execution
Use this operating sequence:
- Ask the agent to invoke
landbot-automation. - Require
RUBE_SEARCH_TOOLSfor the exact task, not a generic query. - Ask it to verify the Landbot connection with
RUBE_MANAGE_CONNECTIONS. - Review the returned tool schema and proposed execution plan.
- Approve only the specific action you want performed.
- Ask for a summary of executed tool calls, affected Landbot objects, and any unresolved fields.
For higher-risk changes, add “read-only discovery first” to your prompt. For bulk changes, ask the agent to process one example object first, then generalize only after validation.
Repository files to read first
The upstream skill is compact: the key file is SKILL.md under composio-skills/landbot-automation. There are no extra rules/, resources/, references/, or helper scripts shown in the repository preview, so most adoption decisions should be based on the skill’s MCP requirements and workflow pattern.
Read SKILL.md for three things: the required MCP server, the Landbot connection check, and the mandatory tool discovery pattern. Also review Composio’s Landbot toolkit documentation at composio.dev/toolkits/landbot if you need to understand which Landbot operations are exposed beyond what the agent discovers at runtime.
landbot-automation skill FAQ
Is landbot-automation enough without Rube MCP?
No. landbot-automation depends on Rube MCP and Composio’s Landbot toolkit. If your client cannot use MCP tools, or if RUBE_SEARCH_TOOLS is unavailable, the skill loses its main execution mechanism. In that case, a normal prompt can still help you plan Landbot changes, but it should not be treated as tool-backed automation.
How is this better than an ordinary Landbot prompt?
An ordinary prompt can brainstorm chatbot flows or suggest API calls. The landbot-automation guide is different because it instructs the agent to discover live tool schemas before acting. That makes it more suitable for operational work where exact fields, tool slugs, connection status, and execution order matter.
Can beginners use this skill safely?
Yes, if they keep the first run read-only. Beginners should ask the agent to discover tools, check connection status, and explain the proposed action before making changes. The main beginner risk is approving a write action without understanding which bot, block, contact, or workspace object will be affected.
When should I not use this skill?
Do not use it when you need purely creative chatbot copy, when your organization forbids third-party automation access, or when you cannot authorize a Landbot connection through Composio. Also avoid using it for high-impact production changes unless you can review the discovered schema and confirm the target objects before execution.
How to Improve landbot-automation skill
Improve landbot-automation results with sharper prompts
The best results come from task-specific prompts. Include the target Landbot object, desired outcome, allowed actions, and confirmation rules.
Example:
“Use landbot-automation to discover tools for exporting or listing Landbot bot metadata. Verify the Landbot connection first. If multiple tools can perform the task, compare them briefly and choose the least destructive option. Do not update, delete, publish, or overwrite anything.”
This gives the agent a precise search use case and prevents it from jumping from discovery into execution.
Common failure modes to prevent
The most common failure is schema guessing. Prevent it by explicitly saying: “Always call RUBE_SEARCH_TOOLS before choosing a tool.” Another failure is inactive authentication; require a RUBE_MANAGE_CONNECTIONS check before any workflow. A third issue is ambiguous object targeting, such as “update the sales bot” when multiple bots may match. Provide exact names, IDs when available, and a rule for handling duplicates.
Iterate after the first output
After the first tool discovery result, ask the agent to restate the schema in practical terms: required fields, optional fields, dangerous actions, and missing information. Then provide the missing values and request a dry-run plan. For write operations, ask for a final confirmation message that includes the tool slug, target object, parameters, and expected effect.
What would make the skill stronger
The upstream skill would be stronger with examples for common Landbot tasks, read-only versus write-operation patterns, and sample prompts for listing bots, updating content, or auditing connections. Until those are added, users should compensate by making their prompt explicit about discovery-first behavior, connection status, approval gates, and rollback expectations.