crustdata-automation
by ComposioHQcrustdata-automation helps Claude run Crustdata workflows through Composio's Rube MCP by discovering current tool schemas first, checking the Crustdata connection, and executing schema-aware calls.
This skill scores 67/100, which means it is acceptable for listing but limited. Directory users get enough evidence to understand that it enables Crustdata automation through Rube MCP and how an agent should begin tool discovery and connection setup, but they should expect a lightweight wrapper rather than a rich, task-specific automation playbook.
- Valid frontmatter with a clear description and explicit MCP requirement: Rube MCP is required and Crustdata tasks should be routed through Composio.
- Prerequisites and setup steps tell agents to verify RUBE_SEARCH_TOOLS, manage the Crustdata connection, and confirm ACTIVE status before execution.
- The skill strongly instructs agents to call RUBE_SEARCH_TOOLS first for current schemas, reducing stale-tool/schema guesswork.
- No support files, scripts, references, or README beyond SKILL.md, so adoption depends entirely on the brief instructions in the skill file.
- Guidance is mostly a generic Rube MCP discovery pattern for Crustdata rather than detailed Crustdata-specific workflows or examples.
Overview of crustdata-automation skill
What crustdata-automation is for
crustdata-automation is a Claude skill for running Crustdata workflows through Composio’s Rube MCP server. It is not a standalone scraper, SDK wrapper, or data model. Its main value is procedural: it tells the agent to discover the current Crustdata tool schemas first, verify the Rube/Crustdata connection, and then execute the task using the right MCP tool rather than guessing parameters from memory.
Best-fit users and jobs
This skill fits users who already use Claude with MCP and want help automating Crustdata operations inside an agent workflow. Typical use cases include asking for company, people, market, or prospecting data from Crustdata and having the agent translate that intent into available Composio Crustdata tool calls. It is most useful when tool schemas may change, because the skill explicitly prioritizes RUBE_SEARCH_TOOLS before execution.
Main differentiator for Workflow Automation
The key difference between the crustdata-automation skill and an ordinary “use Crustdata” prompt is the enforced workflow discipline. The skill centers on discovery, connection validation, and schema-aware execution through Rube MCP. That reduces failed calls caused by stale tool names, missing auth, or invented fields, which are common blockers in Workflow Automation involving external data tools.
How to Use crustdata-automation skill
crustdata-automation install and setup context
Install the skill from the repository path using your skill manager, for example:
npx skills add ComposioHQ/awesome-claude-skills --skill crustdata-automation
Then configure Rube MCP in your client by adding the MCP endpoint:
https://rube.app/mcp
Before expecting useful output, confirm that RUBE_SEARCH_TOOLS is available. Next, use RUBE_MANAGE_CONNECTIONS with toolkit crustdata and complete the returned authorization flow if the connection is not ACTIVE. Do not start data workflows until the Crustdata connection is active.
Inputs the skill needs from you
A weak request is: “Find leads with Crustdata.” A stronger request gives the agent enough detail to search the right tools and choose fields correctly:
Use crustdata-automation to find B2B SaaS companies in the US with 50-500 employees, recent hiring signals, and likely need for data infrastructure. Return company name, website, LinkedIn URL, employee range, signal evidence, and confidence. First discover current Crustdata tools and schemas through Rube MCP, then run the workflow only after confirming the Crustdata connection is active.
Good inputs usually include the target entity type, filters, required output fields, geography, recency expectations, ranking criteria, and acceptable gaps. This matters because the skill depends on live tool discovery; clearer goals produce better RUBE_SEARCH_TOOLS queries and fewer irrelevant execution paths.
Practical crustdata-automation usage flow
Start every run with tool discovery:
RUBE_SEARCH_TOOLS with a use case that matches your exact task, not a generic phrase. For example, use “find recently funded cybersecurity companies and return decision-maker contacts” instead of “Crustdata operations.”
Then check the Crustdata connection through RUBE_MANAGE_CONNECTIONS. If active, ask the agent to follow the returned schema exactly, including required fields and recommended execution plans. After the tool call, have the agent normalize the response into your requested format and flag missing or uncertain fields rather than silently filling them in.
Repository files to read first
The repository for this skill is intentionally small. Read composio-skills/crustdata-automation/SKILL.md first; it contains the operational contract, prerequisites, setup steps, tool discovery rule, and core workflow pattern. There are no extra rules/, resources/, references/, or helper scripts in the provided file tree, so the skill’s reliability depends more on following the MCP workflow than on hidden supporting assets.
crustdata-automation skill FAQ
Do I need Rube MCP and a Crustdata account?
Yes. The crustdata-automation skill requires Rube MCP and an active Crustdata connection through Composio. The skill can guide the agent to check tool availability and connection status, but it cannot bypass authentication or replace access to Crustdata-backed tools.
Why not just prompt Claude to use Crustdata?
A plain prompt may work if the model already has the right tool names and schemas available, but that is fragile. This skill tells the agent to call RUBE_SEARCH_TOOLS first so it can retrieve current tool slugs, input schemas, execution plans, and pitfalls. That makes it better for repeatable Workflow Automation where reliability matters more than a one-off answer.
Is crustdata-automation beginner-friendly?
It is beginner-friendly if your MCP client is already configured. The main concepts to understand are: Rube is the MCP server, Composio provides the Crustdata toolkit, and the agent must discover tools before using them. If you have never configured MCP or connected a toolkit before, expect a short setup step before the skill becomes useful.
When should I not use this skill?
Do not use it when you need offline analysis with no external tool calls, when your client cannot access Rube MCP, or when you need a custom Crustdata integration outside Composio’s toolkit. It is also not the right place to hard-code old schemas; the skill is designed around live discovery precisely because tool definitions can change.
How to Improve crustdata-automation skill
Improve crustdata-automation prompts with constraints
The fastest way to improve results is to add concrete constraints. Specify who or what to search for, which attributes matter, how results should be ranked, and what format you need. For example, add “prioritize companies with verified hiring signals in the last 90 days” or “exclude agencies and marketplaces” instead of asking for a broad company list.
Reduce common failure modes
Most failures come from skipping discovery, running before auth is active, or giving the agent an ambiguous business goal. Tell the agent explicitly: “Search current Crustdata tools first, check the connection, use only returned schemas, and ask me before making assumptions about missing filters.” This keeps the workflow aligned with the skill’s intended guardrails.
Iterate after the first output
After the first run, refine based on observed gaps. If results are too broad, add exclusion criteria. If important fields are missing, ask whether the discovered schema supports them before retrying. If the output is hard to use, request a stable table schema such as company, website, evidence, source field, confidence, and next action.
Extend the skill responsibly
If you customize the skill, preserve the “search tools first” rule. Useful additions would be example prompts for common Crustdata workflows, preferred output schemas, and decision rules for handling missing data. Avoid embedding fixed tool names or parameters unless you also instruct the agent to validate them with RUBE_SEARCH_TOOLS before execution.