datagma-automation

Overview of datagma-automation skill

What datagma-automation does

datagma-automation is a Claude skill for running Datagma workflows through Composio’s Rube MCP server. It is built for lead research and enrichment tasks where the agent needs to discover the current Datagma tool schema, confirm the Datagma connection, and then call the right Rube tool instead of guessing parameters from memory.

The key value is not a long prompt template; it is the enforced workflow: search tools first, verify the Datagma connection, use the returned schema, execute, then validate outputs. That makes the datagma-automation skill useful when Datagma tool names or input fields may change.

Best fit for Lead Research teams

Use datagma-automation for Lead Research when you want an AI agent to help with prospect enrichment, company or contact lookup, data completion, and similar Datagma-backed operations. It is especially relevant for sales ops, growth teams, RevOps, agencies, and founders who already use Claude with MCP and want fewer brittle manual tool calls.

It is not a standalone lead database, scraper, or CRM. The skill assumes the real work happens through Datagma tools exposed by Rube MCP.

Main adoption requirements

Before installing or relying on this skill, confirm three things:

• Your Claude-compatible client can add an MCP server.
• Rube MCP is configured with https://rube.app/mcp.
• A Datagma connection can be activated through RUBE_MANAGE_CONNECTIONS.

The repository contains a single SKILL.md, so there are no helper scripts, examples folder, or local package files to inspect. The install decision mainly depends on whether your environment supports Rube MCP and whether your use case maps to Datagma’s available toolkit actions.

How to Use datagma-automation skill

datagma-automation install and setup path

Install the skill from the source repository with:

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

Then configure Rube MCP in your client by adding:

https://rube.app/mcp

After MCP is available, test that RUBE_SEARCH_TOOLS responds. Next, call RUBE_MANAGE_CONNECTIONS with toolkit datagma. If the connection is not ACTIVE, follow the returned authentication link and re-check status before asking the agent to run Datagma tasks.

Read composio-skills/datagma-automation/SKILL.md first. There are no companion README.md, rules/, resources/, or scripts/ files in this skill folder, so the source skill file is the authoritative implementation guide.

Inputs the skill needs

For reliable datagma-automation usage, give the agent a specific business objective, the known fields, the desired output format, and any limits. Weak prompts like “research this lead” force the model to infer too much. Stronger prompts include the target, what is already known, what counts as a match, and how results should be returned.

Example:

Use datagma-automation for Lead Research. I need to enrich this prospect: name “Jane Smith”, company “Acme Robotics”, domain “acmerobotics.com”. First discover the current Datagma tools with RUBE_SEARCH_TOOLS, verify the Datagma connection is active, then use the appropriate tool schema. Return only fields Datagma provides, include confidence or ambiguity notes, and format the result as a table with source fields and missing fields.

This works better because it tells the agent which workflow to follow, prevents hallucinated fields, and makes validation easier.

Practical workflow to invoke the skill

A good datagma-automation guide workflow is:

1. Ask the agent to call RUBE_SEARCH_TOOLS for the exact Datagma use case.
2. Reuse the returned session ID if available.
3. Check RUBE_MANAGE_CONNECTIONS for toolkit datagma.
4. If active, select the tool slug and parameters from the discovered schema.
5. Execute the Datagma operation through Rube MCP.
6. Review missing, ambiguous, or low-confidence results before using them in outreach or CRM updates.

Do not ask the agent to skip tool discovery. The upstream skill explicitly treats current schemas as mandatory because Rube may return updated tool slugs, required fields, execution plans, or pitfalls.

Prompt patterns that improve output

For lead enrichment, include identifiers in priority order: email, domain, company name, person name, LinkedIn URL, location, and role. For company research, include domain, legal name, country, and any disambiguation clues. If you are processing a list, start with a small batch first so you can inspect field quality and rate-limit behavior before scaling.

Also state what not to do. For example: “Do not invent emails,” “Do not overwrite existing CRM fields unless Datagma returns a clear value,” or “Flag multiple possible matches instead of choosing silently.”

datagma-automation skill FAQ

Is datagma-automation only for Datagma?

Yes. The skill is scoped to Datagma operations exposed through Composio’s Datagma toolkit via Rube MCP. It may be used inside a broader sales or research workflow, but its actionable tool layer is Datagma-specific.

How is this better than an ordinary prompt?

An ordinary prompt can describe a lead research task, but it may not know the current Rube tool schema or whether your Datagma connection is active. The datagma-automation skill gives Claude a repeatable operating pattern: discover tools first, verify connection, then execute with the returned schema. That reduces broken calls and guessed parameters.

Can beginners use this skill?

Beginners can use it if they are comfortable adding an MCP server and following an authentication link. The main learning curve is not Datagma syntax; it is understanding that the agent must call RUBE_SEARCH_TOOLS before execution and should not rely on hard-coded tool names.

When should I not use it?

Do not use datagma-automation if you need offline enrichment, non-Datagma data providers, web scraping outside Datagma, or a fully packaged app with UI, scripts, and stored workflows. Also avoid it when your organization cannot authorize a Datagma connection through Rube MCP.

How to Improve datagma-automation skill

Improve datagma-automation results with better goals

The biggest quality lever is specificity. Replace “find info about this company” with a task such as: “Enrich company domain, industry, employee range, headquarters, and decision-maker contact fields for these five accounts; mark unavailable fields as null; do not infer missing values.”

Clear goals help the agent choose the right Datagma operation after RUBE_SEARCH_TOOLS returns available tools.

Avoid common failure modes

Common issues include inactive Datagma connections, skipped tool discovery, incomplete lead identifiers, and overconfident interpretation of sparse results. If the first output looks wrong, ask the agent to show which Datagma tool slug and input schema it used, then rerun with stronger identifiers or narrower matching rules.

For CRM workflows, separate enrichment from mutation. First retrieve and review enriched data; only then ask for updates in a controlled second step if your toolchain supports it.

Iterate after the first output

After the first run, refine by asking:

• Which records had multiple possible matches?
• Which fields were missing from Datagma rather than merely omitted?
• Which inputs would improve match confidence?
• Should the next batch use the same schema and output columns?

This turns datagma-automation from a one-off lookup into a repeatable Lead Research workflow with clearer acceptance criteria.

Repository improvements worth adding

The skill would be stronger with a short README.md, example prompts for contact and company enrichment, sample RUBE_SEARCH_TOOLS outputs, and troubleshooting notes for inactive connections. A small set of tested prompt recipes would also help users compare ordinary prompting with datagma-automation usage before installing it.