Skip to content
Back to search

osint-skill

SkillWorks with CodexGitHub

OSINT Skill for AI agents (Claude Code, OpenClaw, Codex, OpenCode) — from a name to a scored dossier with psychoprofile, career map, and confidence grades. 55+ Apify actors, 7 search APIs, swarm mode. Early Beta.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/smixs--osint-skill/badge.svg)](/listings/smixs--osint-skill)

HTML

<a href="/listings/smixs--osint-skill"><img src="/api/listings/smixs--osint-skill/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

# Clone the repo
git clone https://github.com/smixs/osint-skill.git
cd osint-skill

# Copy to your agent's skills directory (example for Claude Code)
cp -r osint/ ~/.claude/skills/osint/

# Run self-diagnostics to check what's available
bash osint/scripts/diagnose.sh

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 207 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # OSINT Skill
  2. [![Early Beta](https://img.shields.io/badge/status-early%20beta-orange)](https://github.com/smixs/osint-skill/releases)
  3. [![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)
  4. [![Apify Actors](https://img.shields.io/badge/Apify%20actors-55%2B-green)](https://apify.com)
  5. > **Early Beta** — functional but APIs change, Apify actors rotate, and edge cases are still being discovered.
  6. > **[Версия на русском / Russian version](README.ru.md)**
  7. Systematic intelligence gathering on individuals. From a name or handle to a scored dossier with psychoprofile, career map, and entry points.
  8. ### Compatible Agents
  9. Works with any AI coding agent that supports the SKILL.md format:
  10. | Agent | Install |
  11. |-------|---------|
  12. | **[Claude Code](https://claude.com/product/claude-code)** | `cp -r osint/ ~/.claude/skills/osint/` |
  13. | **[OpenClaw](https://github.com/openclaw/openclaw)** | Copy to workspace `skills/` directory |
  14. | **[Codex](https://developers.openai.com/codex/app)** | Point to `osint/SKILL.md` in your agent config |
  15. | **[OpenCode](https://github.com/anomalyco/opencode)** | Copy to project `skills/` directory |
  16. | **Any SKILL.md agent** | Place `osint/` folder where your agent reads skills |
  17. The skill uses standard tools (`bash`, `curl`, `node`, `python3`) and writes instructions in Markdown — no vendor lock-in.
  18. ## Features
  19. - **Phased pipeline** (0 → 1 → 1.5 → 2 → 3 → 4 → 5 → 6): from quick search to deep research
  20. - **Swarm Mode**: coordinates 3-5 parallel sub-agents on Sonnet for speed
  21. - **55+ Apify actors** embedded: Instagram (12), Facebook (14), TikTok (14), YouTube (5), Google Maps (4), LinkedIn, and more
  22. - **Psychoprofile**: MBTI / Big Five based on content analysis (YouTube transcripts, Telegram messages, blogs)
  23. - **Confidence Scoring**: every fact gets graded A/B/C/D by number of independent confirmations
  24. - **Internal Intelligence**: checks Telegram history, email, vault contacts BEFORE going external
  25. - **Research Escalation**: 4 levels from free to $0.50, from seconds to minutes
  26. - **Budget tracking**: spends ≤$0.50 without asking, asks permission above that
  27. ## Quick Start
  28. ```bash
  29. # Clone the repo
  30. git clone https://github.com/smixs/osint-skill.git
  31. cd osint-skill
  32. # Copy to your agent's skills directory (example for Claude Code)
  33. cp -r osint/ ~/.claude/skills/osint/
  34. # Run self-diagnostics to check what's available
  35. bash osint/scripts/diagnose.sh
  36. ```
  37. ## Requirements
  38. ### Required
  39. | Tool | Purpose | Install |
  40. |------|---------|---------|
  41. | **curl** | HTTP requests to APIs | Pre-installed on macOS/Linux |
  42. | **python3** | JSON parsing, MCP client | Pre-installed on macOS/Linux |
  43. | **jq** | JSON processing | `brew install jq` / `apt install jq` |
  44. ### For Apify actors (55+ platforms)
  45. | Tool | Purpose | Install |
  46. |------|---------|---------|
  47. | **Node.js 18+** | Runs `run_actor.js` (embedded Apify runner) | [nodejs.org](https://nodejs.org) |
  48. ### Optional
  49. | Tool | Purpose | Install |
  50. |------|---------|---------|
  51. | **mcpc** | Dynamic actor discovery in Apify Store | `npm install -g @apify/mcpc` |
  52. ## API Keys & Services
  53. The skill uses **graceful degradation** — the more API keys you provide, the deeper it can dig. You need **at least one** search API to get started.
  54. ### Free Tier
  55. | Service | Env Variable | What It Does | Get It |
  56. |---------|-------------|--------------|--------|
  57. | **Brave Search** | _(built into Claude Code)_ | 2,000 queries/month, basic web search | Built-in, nothing needed |
  58. | **Jina AI** | `JINA_API_KEY` | URL → markdown reader, search, deepsearch | [jina.ai/api-key](https://jina.ai/api-key) |
  59. | **Apify** | `APIFY_API_TOKEN` | Instagram, TikTok, YouTube, LinkedIn scraping. Free tier ~$5/month | [console.apify.com](https://console.apify.com/account/integrations) |
  60. | **Parallel AI** | `PARALLEL_API_KEY` | AI-powered search with reasoning and citations | [platform.parallel.ai](https://platform.parallel.ai) |
  61. ### Paid (recommended)
  62. | Service | Env Variable | What It Does | Cost | Get It |
  63. |---------|-------------|--------------|------|--------|
  64. | **Perplexity API** | `PERPLEXITY_API_KEY` | Sonar (fast AI answers), Deep Research | ~$5/month | [perplexity.ai/settings/api](https://www.perplexity.ai/settings/api) |
  65. | **Exa AI** | `EXA_API_KEY` | Semantic search, people/company research | ~$5/month | [dashboard.exa.ai](https://dashboard.exa.ai) |
  66. | **Tavily** | `TAVILY_API_KEY` | Agent-optimized search, $0.005/request basic | ~$5/month | [app.tavily.com](https://app.tavily.com/home) |
  67. ### Advanced
  68. | Service | Env Variable | What It Does | Cost | Get It |
  69. |---------|-------------|--------------|------|--------|
  70. | **Bright Data** | `BRIGHTDATA_MCP_URL` | CAPTCHA bypass, authwall bypass, Facebook scraping, Yandex search | ~$10/month+ | [brightdata.com/mcp](https://brightdata.com/products/web-scraper/mcp) |
  71. ### Setting Up Keys
  72. **Option 1 — Environment variables (recommended):**
  73. ```bash
  74. export PERPLEXITY_API_KEY="pplx-..."
  75. export EXA_API_KEY="exa-..."
  76. export APIFY_API_TOKEN="apify_api_..."
  77. export JINA_API_KEY="jina_..."
  78. export TAVILY_API_KEY="tvly-..."
  79. export PARALLEL_API_KEY="..."
  80. export BRIGHTDATA_MCP_URL="https://mcp.brightdata.com/..."
  81. ```
  82. **Option 2 — File fallback** (supported by some scripts):
  83. ```
  84. <workspace>/scripts/apify-api-token.txt
  85. <workspace>/scripts/jina-api-key.txt
  86. <workspace>/scripts/parallel-api-key.txt
  87. <workspace>/scripts/brightdata-mcp-url.txt
  88. ```
  89. ## How It Works
  90. ### Research Phases
  91. ```
  92. Phase 0: Tooling Self-Check → diagnose.sh, check available tools
  93. Phase 1: Seed Collection → parallel search across all engines
  94. Phase 1.5: Internal Intelligence → Telegram, email, vault (BEFORE external sources)
  95. Phase 2: Platform Extraction → LinkedIn, Instagram, Facebook, TikTok, YouTube...
  96. Phase 3: Cross-Reference → facts verified, graded A/B/C/D
  97. Phase 4: Psychoprofile → MBTI, Big Five, communication style
  98. Phase 5: Completeness Check → 9 mandatory checks + Depth Score 1-10
  99. Phase 6: Dossier Output → formatted dossier from template
  100. ```
  101. ### Research Escalation (cheap → expensive)
  102. ```
  103. Level 1: Quick Answers → Perplexity Sonar, Brave, Tavily, Exa (~$0.00)
  104. Level 2: Source Verification → Jina read, Parallel extract (~$0.01)
  105. Level 3: Social Media → Apify scrapers, Bright Data (~$0.01-0.10)
  106. Level 4: Deep Research → Perplexity Deep, Exa Deep, Jina Deep (~$0.05-0.50)
  107. ```
  108. ### Embedded Scripts
  109. | Script | Purpose |
  110. |--------|---------|
  111. | `diagnose.sh` | Self-diagnostics for all tools and APIs |
  112. | `perplexity.sh` | search / sonar / deep research |
  113. | `tavily.sh` | search / deep / extract |
  114. | `exa.sh` | search / company / people / crawl / deep |
  115. | `first-volley.sh` | Parallel search across all engines at once |
  116. | `merge-volley.sh` | Deduplicate and group search results |
  117. | `apify.sh` | LinkedIn / Instagram / any actor / store search |
  118. | `run-actor.sh` | Universal Apify runner (55+ actors, polling, CSV/JSON export) |
  119. | `run_actor.js` | Node.js engine powering run-actor.sh |
  120. | `jina.sh` | read URL / search / deepsearch |
  121. | `parallel.sh` | search / extract |
  122. | `brightdata.sh` | scrape / search / search-geo / search-yandex |
  123. | `mcp-client.py` | Lightweight MCP client for Bright Data (stdlib only) |
  124. ## Project Structure
  125. ```
  126. osint/
  127. ├── SKILL.md # Main skill file (452 lines)
  128. ├── references/
  129. │ ├── tools.md # Full catalog of 55+ Apify actors + all tools
  130. │ ├── platforms.md # Platform-specific extraction guide
  131. │ ├── content-extraction.md # YouTube/podcast/blog extraction
  132. │ └── psychoprofile.md # MBTI/Big Five methodology
  133. ├── assets/
  134. │ └── dossier-template.md # Output dossier template
  135. └── scripts/
  136. ├── diagnose.sh # Self-check
  137. ├── run-actor.sh # Universal Apify runner (bash wrapper)
  138. ├── run_actor.js # Apify runner engine (Node.js, embedded)
  139. ├── package.json # ESM support for run_actor.js
  140. ├── apify.sh # Apify shortcuts
  141. ├── perplexity.sh # Perplexity API
  142. ├── tavily.sh # Tavily API
  143. ├── exa.sh # Exa AI API
  144. ├── jina.sh # Jina AI API
  145. ├── parallel.sh # Parallel AI API
  146. ├── brightdata.sh # Bright Data MCP
  147. ├── mcp-client.py # MCP client (Python, stdlib only)
  148. ├── first-volley.sh # Parallel first search
  149. └── merge-volley.sh # Result merging
  150. ```
  151. ## Known Issues (Beta)
  152. - **Shell injection**: user input is interpolated into JSON without `jq` escaping. Do not run with untrusted input.
  153. - **macOS**: `first-volley.sh` uses `tail --pid` (Linux-only). Parallel searches work on macOS, but timeout logic may not trigger.
  154. - **Apify actors**: actor IDs can change or get removed without notice. Use `apify.sh store-search` to find alternatives.
  155. - **Inconsistent key loading**: Perplexity, Tavily, and Exa only load from env vars (no file fallback, unlike Apify/Jina/Parallel).
  156. ## Credits
  157. - **Apify Actor Runner** (`run_actor.js`) embedded from [apify/agent-skills](https://github.com/apify/agent-skills) (MIT License)
  158. - Actor catalog based on [apify-ultimate-scraper](https://github.com/apify/agent-skills/tree/main/skills/apify-ultimate-scraper) v1.3.0
  159. ## License
  160. [MIT](LICENSE)

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.