findymail-automation
by ComposioHQfindymail-automation helps agents run Findymail lead research via Composio Rube MCP by discovering current tool schemas, checking the active connection, and avoiding guessed email results.
This skill scores 64/100, which makes it acceptable but limited for directory listing. Directory users can understand that it is a Rube MCP-based Findymail automation helper and get enough setup guidance to try it, but should expect a thin, generic workflow wrapper rather than a deeply documented, task-specific automation package.
- Valid skill frontmatter clearly declares the Rube MCP dependency and describes the intended Findymail automation scope.
- Prerequisites and setup steps explain how to connect Rube MCP, manage the Findymail connection, and verify ACTIVE status before use.
- The skill repeatedly instructs agents to call RUBE_SEARCH_TOOLS first, which should reduce schema drift and improve triggerability with current tool definitions.
- No support files, examples, scripts, or README are present beyond SKILL.md, so users get limited evidence of tested real-world Findymail workflows.
- The workflow guidance is mostly generic Rube MCP discovery and connection setup rather than detailed task-specific Findymail automations.
Overview of findymail-automation skill
What findymail-automation is for
findymail-automation is a Claude skill for running Findymail lead research and email-finding workflows through Composio’s Rube MCP. Instead of asking an agent to “find emails” generically, the skill forces the agent to discover the current Findymail tool schemas first, check the authenticated connection, and then execute the right Rube tools with valid inputs.
Best fit for Lead Research teams
The findymail-automation skill is best for sales operations, recruiters, founders, growth teams, and lead researchers who already use Findymail or want an AI agent to help enrich prospect lists. It is most useful when you have concrete prospect data such as names, companies, domains, LinkedIn URLs, or target accounts, and you want the agent to automate repeatable lookup steps without guessing API parameters.
What makes this skill different
The main differentiator is its Rube MCP-first workflow. The skill explicitly tells the agent to call RUBE_SEARCH_TOOLS before taking action, because Composio tool names and schemas can change. That makes it safer than a hardcoded prompt that assumes a fixed Findymail API shape. It also includes connection-checking guidance so the agent does not waste time attempting workflows before the Findymail connection is active.
Adoption constraints to know
This is not a standalone scraper or a local CLI. You need an MCP-capable client, Rube MCP configured at https://rube.app/mcp, and an active Findymail connection through Composio. The repository path is composio-skills/findymail-automation, and the main file to inspect is SKILL.md; there are no extra scripts, references, or helper resources in the skill folder.
How to Use findymail-automation skill
findymail-automation install context
Install the skill from the Composio skills repository if your client supports skill installation:
npx skills add ComposioHQ/awesome-claude-skills --skill findymail-automation
Then configure Rube MCP in your AI client by adding the MCP server endpoint:
https://rube.app/mcp
Before running a real lead research task, ask the agent to verify that RUBE_SEARCH_TOOLS is available. Next, use the Rube connection manager for toolkit findymail and complete the returned authentication flow if the connection is not ACTIVE.
Inputs the skill needs
For best findymail-automation usage, provide structured prospect context instead of a vague request. Useful inputs include:
- Person name and company name
- Company domain
- LinkedIn profile URL, if available
- Job title or seniority filter
- Target geography or segment
- Output format, such as CSV columns or a table
- Rules for confidence, exclusions, or manual review
A weak prompt is: “Find emails for these leads.” A stronger prompt is: “Use findymail-automation for Lead Research. First discover current Findymail tools with RUBE_SEARCH_TOOLS, confirm the Findymail connection is active, then enrich this list with work emails where possible. Return name, company, domain, email, confidence, source/tool used, and needs_review. Do not invent emails.”
Practical workflow
A reliable workflow is:
- Read
SKILL.mdto understand the required Rube flow. - Ask the agent to run
RUBE_SEARCH_TOOLSfor the exact Findymail task, such as “find a verified work email from name and domain.” - Ask it to inspect the returned schemas and execution plan before calling any action tool.
- Confirm the Findymail connection is active through the connection-management tool.
- Run a small batch first, review the output, then scale to the full list.
- Require the agent to separate verified results from uncertain or missing results.
This pattern reduces schema errors and prevents the agent from fabricating unavailable fields.
Prompt template you can adapt
Use this prompt when starting:
“Use the findymail-automation skill. My goal is to enrich a lead list with verified work emails. First call RUBE_SEARCH_TOOLS for the current Findymail schemas and recommended execution plan. Check that the Findymail connection is active. Then process the leads below in batches of [batch size]. Return a table with [columns]. Mark any missing, ambiguous, or low-confidence result as needs_review and explain the reason briefly. Do not guess emails or use tools that were not discovered in the current session.”
findymail-automation skill FAQ
Is findymail-automation only for technical users?
No, but it assumes your AI client can use MCP tools. Non-technical users can still benefit if the environment is already configured. The hardest part is not the prompt; it is making sure Rube MCP is available and the Findymail toolkit connection is active.
How is this better than a normal prompt?
A normal prompt may produce plausible but unsupported steps, especially if it assumes outdated tool names. The findymail-automation skill adds operational discipline: discover tools first, inspect schemas, check authentication, and execute only against available Findymail actions. That is valuable for lead research where accuracy and auditability matter.
When should I not use this skill?
Do not use it if you need broad web scraping, personal email guessing, unsupported data collection, or enrichment from tools other than Findymail. It is also a poor fit if you cannot connect Rube MCP or do not have permission to process the prospect data you are providing.
What should I read before installing?
Read composio-skills/findymail-automation/SKILL.md first. It contains the prerequisites, setup path, tool discovery requirement, and core workflow pattern. Because the folder does not include extra scripts or a README, the skill file is the authoritative implementation guide.
How to Improve findymail-automation skill
Improve inputs before improving prompts
The fastest way to improve findymail-automation results is to provide cleaner lead data. Include domains whenever possible, standardize company names, remove duplicates, and separate incomplete records. A person name plus a company domain is usually much more actionable than a name plus a vague company label.
Avoid common failure modes
Common issues include skipping RUBE_SEARCH_TOOLS, running before the Findymail connection is active, processing too many leads before testing, and accepting guessed emails as real results. In your prompt, explicitly require schema discovery, connection verification, batch processing, and a needs_review status for uncertain records.
Iterate after the first output
After the first batch, inspect missing and low-confidence rows. Then ask the agent to adjust the workflow: retry only records with enough input data, normalize domains, or change the requested Findymail tool based on the discovered schema. This makes the findymail-automation guide practical for real lead operations instead of a one-shot enrichment attempt.
Add team-specific guardrails
For production use, add your own rules around allowed regions, consent requirements, CRM field names, deduplication, and what counts as a usable confidence level. The upstream skill gives the Rube and Findymail automation pattern; your team should define the acceptance criteria that decide whether a lead is ready for outreach.