lead-research-assistant
by ComposioHQlead-research-assistant helps agents qualify target companies, rank lead fit, and suggest outreach angles from your ICP, market, and sales goals. Use this single-file skill for structured Lead Research, then verify company and contact data manually.
This skill scores 70/100, which means it is acceptable for listing but best treated as a lightweight, prompt-based workflow rather than a fully tooled lead-research system. Directory users can understand when to invoke it and what output to expect, but should be aware that execution will require available web/search capabilities and some judgment about data sources and lead validation.
- Clear frontmatter and usage triggers for sales, business development, marketing, target-account research, partnerships, and ICP matching.
- The SKILL.md lays out a practical lead-research flow: understand the business, identify target companies, prioritize leads, provide contact strategies, and enrich company/decision-maker context.
- The content appears substantive rather than placeholder/demo material, with multiple workflow-oriented sections and example-style usage guidance.
- Prompt-only skill with no support files, scripts, references, or data-source guidance, so lead quality depends heavily on the agent's available search/tools and user-provided context.
- No install command or repository-level README/metadata is present for this skill path, which slightly weakens adoption clarity for directory users.
Overview of lead-research-assistant skill
What lead-research-assistant does
The lead-research-assistant skill helps an AI agent turn a product description, target market, and sales goal into a structured lead research workflow. It is designed for Lead Research tasks such as identifying target companies, qualifying account fit, ranking opportunities, and suggesting personalized outreach angles.
Best fit for sales and business development users
Use this skill if you need a practical first pass at account discovery for sales, partnerships, agency prospecting, founder-led outreach, or marketing campaign planning. It is most useful when you can describe your offer, ideal customer profile, geography, company size, and the business problem you solve.
What makes the skill useful
The main value of the lead-research-assistant skill is not just “find companies.” It guides the agent to reason through fit: industry, company size, location, tech stack, funding stage, likely pain points, decision-maker context, and outreach strategy. That makes the output more actionable than a generic list of prospects.
Key limitations before installing
This skill has a single SKILL.md file and no bundled scripts, datasets, enrichment APIs, or verification tools. It can structure research and reasoning, but the quality depends on the model’s browsing/tool access, your input detail, and your willingness to verify company and contact data before outreach.
How to Use lead-research-assistant skill
lead-research-assistant install and files to inspect
Install from the ComposioHQ skill collection with:
npx skills add ComposioHQ/awesome-claude-skills --skill lead-research-assistant
After install, read lead-research-assistant/SKILL.md first. There are no extra rules/, resources/, references/, or scripts/ folders in the repository preview, so the skill behavior is concentrated in that one file. This makes setup simple, but also means you should supply your own market context, data sources, and qualification rules.
Inputs the skill needs for better lead research
For strong lead-research-assistant usage, do not ask only for “good leads.” Provide:
- Product or service summary
- Main value proposition and pain solved
- Ideal customer profile
- Excluded customer types
- Target countries or regions
- Preferred company size or revenue range
- Relevant industries
- Buyer titles or departments
- Proof points, case studies, or differentiators
- Outreach objective, such as demo booking, partnership, pilot, or survey
A weak prompt is: “Find SaaS leads for my product.”
A stronger prompt is: “Use lead-research-assistant for a B2B SaaS product that automates SOC 2 evidence collection for 50–500 person fintech companies in the US and UK. Prioritize companies hiring security or compliance roles, using cloud infrastructure, and likely preparing for enterprise sales. Exclude consultancies and companies already selling compliance automation. Return 25 accounts with fit rationale, trigger event, likely buyer title, outreach angle, and confidence score.”
Suggested workflow for lead-research-assistant usage
Start by asking the skill to clarify your ICP before generating leads. Then run lead discovery in batches, review the fit criteria, and only then ask for outreach messaging. A practical sequence is:
- Define ICP and disqualifiers.
- Ask for lead categories and search criteria.
- Generate a ranked account list.
- Request evidence for each account’s fit.
- Add contact strategy and likely buyer personas.
- Export into a CRM-friendly table.
- Manually verify company facts and contacts.
This staged workflow reduces hallucinated prospect lists and makes the research easier to audit.
Output format that improves decisions
Ask for columns that force useful judgment, not just names. Good fields include company, website, industry, location, employee range, fit score, why it fits, pain point, trigger signal, recommended buyer, outreach angle, data to verify, and priority. The data to verify column is especially important because this skill does not include an automatic data validation pipeline.
lead-research-assistant skill FAQ
Is lead-research-assistant better than an ordinary prompt?
Yes, if you want a repeatable lead research structure. A generic prompt may produce a loose list of companies. The lead-research-assistant skill gives the agent a clearer job: understand the business, identify matching accounts, prioritize fit, enrich context, and suggest contact strategy.
Does it find verified email addresses or phone numbers?
Not by itself. The repository evidence shows no bundled enrichment scripts or contact database integration. If your environment has browsing, CRM, Apollo, Clay, Clearbit, LinkedIn, or other tools, you can pair those with the skill. Otherwise, treat contact details as suggestions to verify manually.
Who should not use this skill?
Do not rely on it as a compliance-safe prospecting database, a guaranteed email finder, or a replacement for human review. It is also a poor fit if you cannot explain your offer or ICP. Without clear constraints, the output may be broad, generic, or difficult to prioritize.
Is this beginner-friendly?
Yes. The skill is simple to install and centered in SKILL.md, so beginners can adopt it quickly. The main learning curve is prompt quality: the more precisely you describe your market, exclusions, and sales motion, the better the lead list and outreach strategy will be.
How to Improve lead-research-assistant skill
Improve lead-research-assistant prompts with sharper ICPs
The fastest way to improve lead-research-assistant results is to provide hard qualification rules. Replace vague traits like “growing companies” with observable signals such as “raised Series A or B in the last 24 months,” “hiring revenue operations roles,” “uses Salesforce,” or “has multiple compliance job postings.” Observable signals make the lead research easier to validate.
Reduce common failure modes
Common problems include overly broad industries, invented company details, weak prioritization, and outreach angles that sound generic. Counter these by asking the agent to separate confirmed facts from assumptions, cite what should be verified, and explain why each account is a fit. If the first list is noisy, revise the disqualifiers before asking for more leads.
Add your own research sources and scoring model
Because the skill has no bundled data sources, improve it by naming the sources your team trusts: company websites, job boards, funding databases, app marketplaces, review sites, CRM exports, conference sponsor lists, or technology directories. You can also define a scoring model, for example: 40% pain fit, 25% company size, 20% buying trigger, 15% reachability.
Iterate from account list to outreach
After the first output, do not immediately send emails. Ask for a second pass: “Remove weak-fit accounts, group the rest by pain point, and write one personalized outreach hypothesis per segment.” Then request message drafts only for the highest-priority accounts. This keeps the lead-research-assistant for Lead Research workflow focused on qualified opportunities instead of volume.