Introduction

For people doing B2B sales or market research, there is a repetitive workflow that occurs daily: circling target companies by industry and region, verifying information across multiple channels, and organizing the contact information found into a lead sheet. This type of work has clear rules and fixed steps, making it suitable to be handed over to an agent to execute.

The philosophy of DSH (DeepSeek Harness) is that “everything is a plugin.” The community has encapsulated this workflow into a skill called BigLead by user kobenfang. Below is an introduction to its positioning, features, installation methods, and considerations.

What is it

BigLead (repository: kobenfang/biglead) is a skill for B2B lead generation. Its core consists of four things: searching for target companies, cross-verification, extracting contact information, and managing the lead database. The corresponding npm package is @kobenfang/dsh-biglead, version 1.0.0, with an MIT license.

It follows the Agent Skills open standard (SKILL.md + YAML frontmatter), so it is not limited to dsh; frameworks like OpenClaw and Claude Code that support this standard can also load it. The author has also published it to ClawHub for OpenClaw users.

Core Features

The capabilities listed in the README include:

  • Searching for target companies by industry/product/region
  • Cross-verification across multiple channels
  • Extracting contact information (if available)
  • Managing the customer lead database

Applicable scenarios cover B2B sales, market research, and competitive analysis. Two points need clarification first: First, “extract contact information” is marked as “if available,” meaning it is not guaranteed to get it every time; Second, more complete functional descriptions are in the repository’s SKILL.md, and it is recommended to read it before use.

Environment Dependencies

Two conditions: Python 3, and web search capability (web search tool). Both search and cross-verification rely on the latter, so it must be available in the runtime environment.

Installation and Activation

Installation involves placing the repository in the skills directory. The README provides two methods:

# Method 1: Place in user-level shared directory (recommended)
mkdir -p ~/.agents
git clone https://github.com/kobenfang/biglead.git ~/.agents/biglead

# Method 2: Place in project directory
mkdir -p .agents
cp -r biglead .agents/

Method 1 is under the user’s home directory and can be shared by multiple projects; Method 2 only applies to the current project. After cloning or copying, start dsh, and you will find biglead in the skills list.

The README does not provide a native installation command like dsh plugin add; the two methods above are the official installation methods provided. For OpenClaw users via ClawHub:

npx clawhub install biglead

Before installing, you can check using search:

npx clawhub search biglead

Directory Structure

The repository is lightweight, with the core being two files:

biglead/
├── SKILL.md          # Skill definition (Agent Skills standard)
└── scripts/biglead.py

SKILL.md is the skill definition with YAML frontmatter; the complete functional description is also here. scripts/biglead.py is the script implementation. After installation, if you want to confirm what it can do, look at SKILL.md first.

Search Keywords

The search keywords provided in the README are: customer leads, sales, B2B, lead generation, market research, lead generation, sales prospecting. Searching for these terms in the skill list or community directory can help locate this skill.

Applicable Scenarios and Precautions

Suitable groups:

  • B2B sales personnel who need to bulk define target companies by industry and region
  • People doing market research and competitive analysis who need to cross-verify company information

Before enabling, several points to note:

  1. The skill runs with the permissions of the current dsh process; the resources it can access are exactly what you can access. It is recommended to read SKILL.md and scripts/biglead.py thoroughly before installing to confirm there are no issues.
  2. The license is MIT; see the repository for specific terms.
  3. Contact information extraction is “if available”; whether it is obtained depends on whether it exists on public channels. Do not treat “contact information is guaranteed” as an expectation.
  4. If you found it through a community directory: The directory site is an independent community project with no official affiliation with DeepSeek or Matrix, and the plugin itself is maintained by community members.

Conclusion

BigLead turns the fixed workflow of “search company — verify — extract contact — manage leads” in B2B lead generation into a lightweight skill; installation only requires one git clone. If you frequently do customer development or market research, you can install it and try it.

  • Community Directory Page: https://www.skillhub.cn/plugins/kobenfang/BigLead
  • GitHub Repository: https://github.com/kobenfang/BigLead