rafflys-automation
by ComposioHQrafflys-automation helps agents automate Rafflys operations through Composio Rube MCP by discovering live tool schemas, checking the Rafflys connection, and running workflows safely.
This skill scores 66/100, which makes it acceptable but limited for directory listing. Directory users can understand when to use it and how to start a Rafflys connection through Rube MCP, but they should expect a lightweight connector-style skill rather than a rich Rafflys workflow pack with concrete task recipes.
- Clear trigger and scope: it is explicitly for automating Rafflys operations through Composio's Rafflys toolkit via Rube MCP.
- Operational prerequisites are stated, including Rube MCP availability, an active Rafflys connection, and use of RUBE_SEARCH_TOOLS before execution.
- Includes a repeatable discovery-first pattern using RUBE_SEARCH_TOOLS and RUBE_MANAGE_CONNECTIONS, reducing schema guesswork for agents.
- No support files, scripts, references, or examples beyond SKILL.md, so adoption depends entirely on the short written instructions.
- Workflow guidance appears mostly generic to Rube MCP and does not show concrete Rafflys task examples or Rafflys-specific fields/actions in the provided evidence.
Overview of rafflys-automation skill
What rafflys-automation is for
rafflys-automation is a Claude skill for running Rafflys operations through Composio’s Rube MCP server. It is designed for users who want an AI agent to discover the current Rafflys tool schemas, verify the Rafflys connection, and then execute task-specific workflows with less guesswork than a generic prompt.
The key value is not a large local codebase; the skill is mainly an operating procedure. It tells the agent to use RUBE_SEARCH_TOOLS first, because Rafflys tool names and input schemas should be treated as live MCP data rather than hard-coded assumptions.
Best-fit users and workflows
This rafflys-automation skill is a good fit if you already use Claude or another MCP-capable client and need to automate Rafflys work through Composio. Typical use cases include asking the agent to inspect available Rafflys actions, prepare a workflow from the returned schema, check authentication state, and execute a specific Rafflys operation only after confirming required fields.
It is especially useful for Workflow Automation teams that prefer tool-discovery-first execution rather than brittle prompt instructions copied from old docs.
What makes it different from a normal prompt
A normal prompt may say “use Rafflys,” but it may invent tool names, miss connection setup, or use stale parameters. rafflys-automation adds a repeatable sequence: connect Rube MCP, manage the Rafflys connection, search current tools, review returned schemas, then run the selected tool. That sequence is the main adoption reason.
Important adoption constraints
The skill depends on Rube MCP and an active Rafflys connection. It does not ship helper scripts, local resources, or extra rules folders; the central file is SKILL.md. If your environment cannot add https://rube.app/mcp as an MCP server, or if you cannot authorize the Rafflys toolkit through RUBE_MANAGE_CONNECTIONS, this skill will not be useful yet.
How to Use rafflys-automation skill
rafflys-automation install context
Install the skill from the Composio skill collection:
npx skills add ComposioHQ/awesome-claude-skills --skill rafflys-automation
Then configure Rube MCP in your MCP-capable client by adding:
https://rube.app/mcp
The upstream skill notes that no separate API key is required for the MCP endpoint itself, but you still need an active Rafflys connection. Ask your agent to verify RUBE_SEARCH_TOOLS is available before attempting Rafflys work.
First files and tools to inspect
Start with composio-skills/rafflys-automation/SKILL.md. There are no extra scripts/, resources/, references/, or rules/ directories in the repository preview, so the important implementation guidance is in that one file.
The agent should use this order:
- Call
RUBE_SEARCH_TOOLSfor the specific Rafflys task. - Call
RUBE_MANAGE_CONNECTIONSwith toolkitrafflys. - If the connection is not
ACTIVE, complete the returned authorization flow. - Re-run tool discovery if needed and use the latest schema for execution.
Strong prompt pattern for rafflys-automation usage
Weak prompt:
“Automate my Rafflys task.”
Stronger prompt:
“Use the rafflys-automation skill. First confirm Rube MCP is available. Search Rafflys tools for this use case: [describe the exact Rafflys outcome]. Check the rafflys connection status and stop if it is not ACTIVE. Before executing, summarize the selected tool slug, required fields, optional fields, and any risky assumptions. Ask me for missing required values instead of guessing.”
This works better because the agent gets the task goal, the required discovery step, the connection boundary, and a clear rule for missing inputs.
Practical workflow tips
Be specific about the Rafflys object, action, filters, dates, campaign names, user identifiers, or output format you expect. If the tool schema returns required fields you did not provide, treat that as a clarification step, not an error. For safer automation, ask the agent to show the planned call before execution when the action could create, update, or delete Rafflys data.
For repeat workflows, keep the same Rube session ID when appropriate so tool discovery and execution context remain connected.
rafflys-automation skill FAQ
Is rafflys-automation beginner-friendly?
It is beginner-friendly if your client already supports MCP tools. The skill gives a clear setup path: add Rube MCP, verify RUBE_SEARCH_TOOLS, manage the Rafflys connection, then discover tools. Beginners may still need help completing the Rafflys authorization flow returned by RUBE_MANAGE_CONNECTIONS.
When should I not use rafflys-automation?
Do not use it when you need offline automation, local-only scripts, or a fixed Rafflys API wrapper. This skill is built around live Composio/Rube tool discovery. It is also a poor fit if your organization does not allow MCP server connections or third-party workflow tooling.
How does it compare with direct Rafflys API work?
Direct API work may be better for production code where you need versioned endpoints, tests, and strict deployment controls. rafflys-automation is better for agent-driven operations where the assistant can inspect current schemas and perform tasks interactively through Composio’s Rafflys toolkit.
Does rafflys-automation for Workflow Automation require coding?
Usually no. The skill is prompt-and-tool driven. You need to configure MCP and authorize the Rafflys toolkit, but routine usage is mostly about describing the workflow clearly and letting the agent discover and call the right tools.
How to Improve rafflys-automation skill
Improve rafflys-automation inputs
The fastest way to improve results is to replace vague goals with operational details. Include the Rafflys task type, target records, constraints, time range, success condition, and whether the agent may execute immediately or must ask for approval.
Example:
“Find Rafflys tools for exporting entries from campaign X between May 1 and May 31. Do not execute until you show me the required schema fields and confirm which fields are missing.”
Prevent common failure modes
The main failure modes are skipped tool discovery, inactive connection, stale assumptions about tool parameters, and missing required fields. Counter them directly in your prompt: require RUBE_SEARCH_TOOLS first, require a connection check, and instruct the agent to ask questions instead of fabricating values.
If a call fails, ask the agent to compare the attempted payload against the latest returned schema before retrying.
Iterate after the first output
After the first response, refine by narrowing scope: “only active campaigns,” “return a CSV-ready summary,” “exclude test records,” or “dry-run the update plan first.” For destructive or high-impact operations, add an approval checkpoint between schema discovery and execution.
What repository improvements would help
The upstream skill would be stronger with more example Rafflys use cases, sample RUBE_SEARCH_TOOLS queries, and safe execution patterns for create/update/delete actions. Until then, treat SKILL.md as the authoritative guide and make your own prompts explicit about schema discovery, connection status, and approval boundaries.