Lead Research Skill: Verified Prospect Intel for Sales

Where this fits

You are building a sales development agent that needs to clean and enrich a list of 200 potential buyers. Doing it with separate API calls for LinkedIn, email verification, and company data creates a fragile pipeline. The lead-research skill wraps those capabilities into a repeatable research workflow, so an agent can go from raw name/company to a sales-ready lead profile without hand-written glue code.

Why agents benefit

• It defines a clear sequence: understand the target, call sandbase_describe_tool to confirm parameters, then execute the selected tool. The agent doesn't have to guess schemas.
• It combines multiple data sources: LinkedIn profile, email verification, company enrichment, and social presence. The agent can cross-check signals rather than trusting one source.
• It produces structured prospect intelligence, not just a contact record. An agent can use the output to personalize outreach, segment lists, or update CRM fields.
• It reduces invalid outreach by including email verification as a first-class step, not an afterthought.
• It gives the agent a reusable pattern for other research skills: describe before call, use only current input schema, and map results to the user's task.

Practical scenarios

Outbound prospecting for a niche SaaS product

The agent receives target titles like Head of Growth at Series B fintech startups. It enriches company data to filter by size and funding, pulls LinkedIn profiles for decision makers, verifies email addresses, and outputs a CSV ready for an outreach sequence.

Refreshing a stale CRM before a campaign

Many CRM contacts have outdated titles and bounced emails. The agent runs the skill over the list, checks social signals, updates roles and companies, and flags records where the lead has changed jobs or gone dark.

Personalizing warm introductions

Before drafting a message to a referral, the agent researches the person's recent LinkedIn activity and company context. The final brief includes current role, mutual points, and verified contact details, making the agent's ask specific and credible.

Add it to your agent workflow

Install the skill with:

npx skills add sandbaseai/sandbase-skills --skill lead-research

Then in agent code, route through SandBase's two-step interface:

sandbase_describe_tool(tool_name='apollo_company_enrich')
sandbase_call_tool(
    tool_name='apollo_company_enrich',
    arguments={'domain': 'example.com'}
)

Always read the skill's API map first. It tells you which tool fits the user's research question before you start calling capabilities.

Compare before adopting

Before replacing an existing research stack, compare the skill against what you already have: How current are the underlying data sources? Does the skill's workflow fit your agent's model context size and permitted tools? Is a custom script with your own API keys simpler? Also compare against alternatives like direct Apollo, LinkedIn, and NeverBounce integrations. The skill is useful when you want a ready-made researcher rather than another vendor SDK.

Why it is worth tracking

The project has 68 stars and an active GitHub repository from sandbaseai, which is a reasonable signal for a focused agent skill. Since lead data freshness degrades quickly, check whether the skill's maintainers update the API map when providers change schemas. It is worth evaluating now if you are building outbound sales agents, and watching if you already have a working research pipeline. You can inspect the skill's repository and community signals at https://www.openagentskill.com/skills/sandbaseai-lead-research.