Skip to content

CompanyOSINT AgentFlow

github-actions[bot] edited this page Nov 15, 2026 · 1 revision

๐ŸŽฏ Purpose

AnythingLLM AgentFlow for automated company-level OSINT - aggregates public information about a target organization including tech stack, employee data, infrastructure, and social presence.

โš™๏ธ Function

Accepts a company name or domain, queries public OSINT sources, and returns structured intelligence: employee count/structure, technology stack (job postings, LinkedIn), public IP ranges, email formats, social media presence, and known subsidiaries.

๐Ÿ† Goal

Build a comprehensive target profile for a company during the reconnaissance phase of an authorized engagement, without active probing.

๐Ÿ“‹ When to Use

Early-stage reconnaissance on an authorized target organization; scoping an engagement; building the target profile before active testing begins.

Flow Information

Name: CompanyOSINT

Description:

Gathers public information about a target company including employees, email formats, 
technologies used, and social media presence for pre-engagement reconnaissance.

Purpose: Pre-engagement intelligence gathering for penetration testing


Flow Variables

Define these variables in the "Flow Variables" section:

Variable Name Initial Value Description
company_name (empty) Target company name
domain (empty) Company domain (e.g., acme.com)
linkedin_data (empty) LinkedIn scraping results
github_data (empty) GitHub search results
tech_stack (empty) Technology stack information
employee_info (empty) Analyzed employee data
tech_analysis (empty) Technology analysis results
osint_report (empty) Final compiled report

Flow Blocks

Block 1: Web Scraping - LinkedIn

Block Type: Web Scraping

Configuration:

  • URL: https://www.linkedin.com/search/results/people/?keywords=${company_name}
  • Capture Page Content As: Text content only
  • Content Summarization: Enabled (recommended)
  • Result Variable: linkedin_data

Purpose: Gather employee information from LinkedIn


Block 2: Web Scraping - GitHub

Block Type: Web Scraping

Configuration:

  • URL: https://github.com/search?q=${domain}&type=code
  • Capture Page Content As: Text content only
  • Content Summarization: Enabled (recommended)
  • Result Variable: github_data

Purpose: Find company code repositories and exposed information


Block 3: Web Scraping - BuiltWith

Block Type: Web Scraping

Configuration:

  • URL: https://builtwith.com/${domain}
  • Capture Page Content As: Text content only
  • Content Summarization: Enabled (recommended)
  • Result Variable: tech_stack

Purpose: Identify technologies used by the company


Block 4: LLM Instruction - Analyze Employee Info

Block Type: LLM Instruction

Instruction:

Analyze this LinkedIn data for ${company_name}:

${linkedin_data}

Extract and infer:
1. Common employee naming patterns (firstname.lastname@domain)
2. Email format (first.last, flast, firstl, etc.)
3. Key personnel (IT staff, security team, executives)
4. Department structure
5. Technologies mentioned in profiles
6. Potential social engineering targets

Format as structured intelligence report with clear sections.

Result Variable: employee_info

Purpose: Extract actionable intelligence from LinkedIn data


Block 5: LLM Instruction - Technology Analysis

Block Type: LLM Instruction

Instruction:

Analyze technology stack for ${domain}:

BuiltWith Data: ${tech_stack}
GitHub Data: ${github_data}

Identify:
1. Web technologies (frameworks, CMS, servers)
2. Cloud providers and services
3. Third-party integrations
4. Development languages and frameworks
5. Potential attack surfaces
6. Known vulnerabilities in detected technologies

Prioritize findings by security relevance.

Result Variable: tech_analysis

Purpose: Identify technology stack and potential vulnerabilities


Block 6: LLM Instruction - Compile OSINT Report

Block Type: LLM Instruction

Instruction:

Create comprehensive OSINT reconnaissance report:

Company: ${company_name}
Domain: ${domain}

Employee Intelligence: ${employee_info}
Technology Stack: ${tech_analysis}

Format as professional intelligence report:

# OSINT Report: ${company_name}

## Executive Summary
[Brief overview of findings]

## Employee Information
[Employee details, email formats, key personnel]

## Technology Stack
[Technologies, versions, vendors identified]

## Attack Surface
[Potential entry points and vulnerabilities]

## Social Engineering Opportunities
[Potential targets and approaches]

## Recommendations
[Suggested next steps for penetration test]

Use professional security assessment language.

Result Variable: osint_report

Purpose: Compile all findings into professional report


Block 7: Flow Complete

Block Type: Flow Complete

Configuration:

  • Flow will end here and return results

Usage Instructions

Basic Usage

@agent use CompanyOSINT flow
company_name: Acme Corporation
domain: acme.com

Example With Real Data

@agent use CompanyOSINT flow
company_name: Pacific Northwest Computers
domain: pnwcomputers.com

Expected Output

The flow will return osint_report containing:

  • Employee naming conventions and email formats
  • Key personnel identification
  • Technology stack analysis
  • Attack surface assessment
  • Social engineering opportunities
  • Recommended next steps

Integration with Other Flows

Typical Workflow:

  1. CompanyOSINT (this flow) - Gather initial intelligence
  2. DomainRecon - Map technical infrastructure
  3. EmailOSINT - Investigate key personnel emails
  4. GitHubSecrets - Deep dive into code repositories
  5. BreachChecker - Check for compromised accounts

Notes and Considerations

Legal and Ethical

  • โš ๏ธ Only use with proper authorization
  • Ensure activities are within scope of engagement
  • Follow all applicable laws and regulations
  • LinkedIn scraping may violate Terms of Service - use cautiously

Technical Considerations

  • LinkedIn may require authentication or block scraping attempts
  • BuiltWith free tier has limitations
  • GitHub search has rate limits
  • Results quality depends on target's public presence

Optimization Tips

  1. For Better Results:

    • Use full company legal name
    • Try variations if initial search fails
    • Allow time for web scraping (30-60 seconds per site)
  2. Alternative Data Sources:

    • If LinkedIn fails, try company website "About" or "Team" pages
    • Check Crunchbase, AngelList for startup information
    • Search for company in security forums or breach databases
  3. Variable Substitution:

    • Ensure variables use ${variable_name} syntax
    • Test with known companies first
    • If scraping fails, check URLs manually

Troubleshooting

Issue Solution
LinkedIn returns no data Try without authentication or use Google search instead
GitHub search fails Use GitHub API with authentication token
BuiltWith blocks request Add delay or use Wappalyzer alternative
Generic analysis Provide more context in company_name
Variable not substituting Check syntax: ${variable} not {{variable}}

Version History

  • v1.0 - Initial flow creation
  • Purpose: Pre-engagement OSINT for penetration testing
  • Last Updated: November 2025

Related Flows

  • DomainRecon - Technical infrastructure mapping
  • EmailOSINT - Email address investigation
  • GitHubSecrets - Repository credential scanning
  • BreachChecker - Breach data verification

Related Files


๐Ÿ“„ Source: AI/AnythingLLM/flow-company_osint.md | โœ๏ธ Suggest an edit | โฌ…๏ธ Back to Artificial Intelligence

Clone this wiki locally