Skip to content
Back to search

robin

AI-Powered Dark Web OSINT Tool

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/apurvsinghgautam--robin/badge.svg)](/listings/apurvsinghgautam--robin)

HTML

<a href="/listings/apurvsinghgautam--robin"><img src="/api/listings/apurvsinghgautam--robin/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.

> [!NOTE] > The tool needs Tor to do the searches. You can install Tor using `apt install tor` on Linux/Windows(WSL) or `brew install tor` on Mac. Once installed, confirm if Tor is running in the background.

> [!TIP] > Provide your API key either in a `.env` file (copy [`.env.example`](.env.example)) or as environment variables. One key is enough: Robin lists models for whichever providers it finds. Supported keys are `OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, `GOOGLE_API_KEY`, `MISTRAL_API_KEY` and `OPENROUTER_API_KEY`. > > For Ollama, nothing goes in your `.env`. Robin defaults to `http://host.docker.internal:11434`, which is what the recommended Docker install needs. Two things are still on you: > > 1. Run the container with `--add-host=host.docker.internal:host-gateway` (the README command already does). > 2. Make Ollama listen on all interfaces, since it binds to `127.0.0.1` by default and a container cannot reach that. If you start it yourself, `OLLAMA_HOST=0.0.0.0 ollama serve &`. If it runs under systemd, `sudo systemctl edit ollama.service`, add `[Service]` and `Environment="OLLAMA_HOST=0.0.0.0"`, then `sudo systemctl daemon-reload && sudo systemctl restart ollama`. > > See [TROUBLESHOOTING.md](TROUBLESHOOTING.md) if it still doesn't appear. > > For any other OpenAI-compatible provider (LM Studio, llama.cpp, Groq, etc.), use the **πŸ”Œ Custom API Provider** expander in the sidebar β€” no `.env` changes required. Enter the base URL, an optional API key, and optionally a model name if the provider doesn't expose `/v1/models` for auto-discovery.

Docker [Recommended]

- Pull the latest Robin docker image

docker pull apurvsg/robin:latest

- Create a `.env` file in the folder you run from, with your API key in it (see [`.env.example`](.env.example)). Create it before the first run: if it does not exist, Docker mounts an empty folder in its place and Robin starts with no keys.

touch .env   # then add your API key to it

- Run the docker image as:

docker run --rm \
   -v "$(pwd)/.env:/app/.env" \
   --add-host=host.docker.internal:host-gateway \
   -p 8501:8501 \
   apurvsg/robin:latest

> [!TIP] > To persist saved investigations across Docker restarts, mount a named volume or a local directory at `/app/investigations`. > > A named volume works the same on every OS and needs no host path. It is also the volume the agent commands in [Robin as MCP](#robin-as-mcp) mount, so the UI shows the same reports your agent saved: > ```bash > docker run --rm \ > -v "$(pwd)/.env:/app/.env" \ > -v robin-investigations:/app/investigations \ > --add-host=host.docker.internal:host-gateway \ > -p 8501:8501 \ > apurvsg/robin:latest > ``` > Or mount a local directory, if you want the JSON files on your own disk: > ```bash > docker run --rm \ > -v "$(pwd)/.env:/app/.env" \ > -v "$(pwd)/investigations:/app/investigations" \ > --add-host=host.docker.internal:host-gateway \ > -p 8501:8501 \ > apurvsg/robin:latest > ``` > Investigations are saved to the `investigations/` folder in your working directory and can be loaded from the **Past Investigations** panel in the sidebar. > > Create the folder yourself first (`mkdir -p investigations`) so Docker does not create it as root. The container runs as UID 1000, and on Linux a folder owned by anyone else is read-only to it β€” Robin prints one warning and the save fails, though the investigation still runs. If you hit that, or your own UID isn't 1000, use the named volume above, or see [Saved investigations fail with permission denied on Linux](TROUBLESHOOTING.md#saved-investigations-fail-with-permission-denied-on-linux).

- Open your browser and navigate to `http://localhost:8501`

Build it yourself

To run your own build instead of the published image, clone the repository and build it:

docker build -t robin .

Then use the same run commands with `robin` in place of `apurvsg/robin:latest`:

docker run --rm \

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 Β· README Β· 267 lines

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

  1. <div align="center">
  2. <img src=".github/assets/logo.png" alt="Logo" width="300">
  3. <br><a href="https://github.com/apurvsinghgautam/robin/actions/workflows/release.yml"><img alt="Release" src="https://github.com/apurvsinghgautam/robin/actions/workflows/release.yml/badge.svg"></a> <a href="https://github.com/apurvsinghgautam/robin/releases"><img alt="GitHub Release" src="https://img.shields.io/github/v/release/apurvsinghgautam/robin"></a> <a href="https://hub.docker.com/r/apurvsg/robin"><img alt="Docker Pulls" src="https://img.shields.io/docker/pulls/apurvsg/robin"></a>
  4. <p align="center">
  5. <a href="https://www.star-history.com/apurvsinghgautam/robin"><img src="https://api.star-history.com/badge?repo=apurvsinghgautam/robin&type=rank&theme=dark" alt="Star History Rank" /> <img src="https://api.star-history.com/badge?repo=apurvsinghgautam/robin&type=trending&theme=dark" alt="GitHub Trending Repository of the Day" /></a>
  6. </p>
  7. <h1>Robin: AI-Powered Dark Web OSINT Tool</h1>
  8. <p>Robin is an AI-powered tool for conducting dark web OSINT investigations. It leverages LLMs to refine queries, filter search results from dark web search engines, and provide an investigation summary.</p>
  9. <a href="#installation">Installation</a> &bull; <a href="#robin-as-mcp">Robin as MCP</a> &bull; <a href="TROUBLESHOOTING.md">Troubleshooting</a> &bull; <a href="CONTRIBUTING.md">Contributing</a> &bull; <a href="#acknowledgements">Acknowledgements</a><br><br>
  10. </div>
  11. ![Demo](.github/assets/screen-ui.png)
  12. ## Architecture
  13. ![Workflow](.github/assets/robin-workflow.png)
  14. ---
  15. ## Features
  16. - πŸ€– **Multi-Model Support** – OpenAI, Claude, Gemini, Mistral, OpenRouter, Ollama, or any OpenAI-compatible API (LM Studio, llama.cpp, Groq, etc.).
  17. - 🌐 **Web UI** – Streamlit-based interface for interactive investigations.
  18. - πŸ”Œ **MCP Support** – Run Robin from any MCP-capable agent, using that agent's own model. See [Robin as MCP](#robin-as-mcp).
  19. - πŸ’¬ **Conversational Follow-ups** – Ask grounded follow-up questions about an investigation without re-running the search, answered from that investigation's own data.
  20. - πŸ”€ **One-Click Pivots** – Suggested follow-up queries surfaced from the findings; click one to launch a fresh investigation.
  21. - 🐳 **Docker-Ready** – Recommended Docker deployment for clean, isolated usage.
  22. ---
  23. ## ⚠️ Disclaimer
  24. > This tool is intended for educational and lawful investigative purposes only. Accessing or interacting with certain dark web content may be illegal depending on your jurisdiction. The author is not responsible for any misuse of this tool or the data gathered using it.
  25. >
  26. > Use responsibly and at your own risk. Ensure you comply with all relevant laws and institutional policies before conducting OSINT investigations.
  27. >
  28. > Additionally, Robin leverages third-party APIs (including LLMs). Be cautious when sending potentially sensitive queries, and review the terms of service for any API or model provider you use.
  29. ## Installation
  30. > [!NOTE]
  31. > The tool needs Tor to do the searches. You can install Tor using `apt install tor` on Linux/Windows(WSL) or `brew install tor` on Mac. Once installed, confirm if Tor is running in the background.
  32. > [!TIP]
  33. > Provide your API key either in a `.env` file (copy [`.env.example`](.env.example)) or as environment variables. One key is enough: Robin lists models for whichever providers it finds. Supported keys are `OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, `GOOGLE_API_KEY`, `MISTRAL_API_KEY` and `OPENROUTER_API_KEY`.
  34. >
  35. > For Ollama, nothing goes in your `.env`. Robin defaults to `http://host.docker.internal:11434`, which is what the recommended Docker install needs. Two things are still on you:
  36. >
  37. > 1. Run the container with `--add-host=host.docker.internal:host-gateway` (the README command already does).
  38. > 2. Make Ollama listen on all interfaces, since it binds to `127.0.0.1` by default and a container cannot reach that. If you start it yourself, `OLLAMA_HOST=0.0.0.0 ollama serve &`. If it runs under systemd, `sudo systemctl edit ollama.service`, add `[Service]` and `Environment="OLLAMA_HOST=0.0.0.0"`, then `sudo systemctl daemon-reload && sudo systemctl restart ollama`.
  39. >
  40. > See [TROUBLESHOOTING.md](TROUBLESHOOTING.md) if it still doesn't appear.
  41. >
  42. > For any other OpenAI-compatible provider (LM Studio, llama.cpp, Groq, etc.), use the **πŸ”Œ Custom API Provider** expander in the sidebar β€” no `.env` changes required. Enter the base URL, an optional API key, and optionally a model name if the provider doesn't expose `/v1/models` for auto-discovery.
  43. ### Docker [Recommended]
  44. - Pull the latest Robin docker image
  45. ```bash
  46. docker pull apurvsg/robin:latest
  47. ```
  48. - Create a `.env` file in the folder you run from, with your API key in it (see [`.env.example`](.env.example)). Create it before the first run: if it does not exist, Docker mounts an empty folder in its place and Robin starts with no keys.
  49. ```bash
  50. touch .env # then add your API key to it
  51. ```
  52. - Run the docker image as:
  53. ```bash
  54. docker run --rm \
  55. -v "$(pwd)/.env:/app/.env" \
  56. --add-host=host.docker.internal:host-gateway \
  57. -p 8501:8501 \
  58. apurvsg/robin:latest
  59. ```
  60. > [!TIP]
  61. > To persist saved investigations across Docker restarts, mount a named volume or a local directory at `/app/investigations`.
  62. >
  63. > A named volume works the same on every OS and needs no host path. It is also the volume the agent commands in [Robin as MCP](#robin-as-mcp) mount, so the UI shows the same reports your agent saved:
  64. > ```bash
  65. > docker run --rm \
  66. > -v "$(pwd)/.env:/app/.env" \
  67. > -v robin-investigations:/app/investigations \
  68. > --add-host=host.docker.internal:host-gateway \
  69. > -p 8501:8501 \
  70. > apurvsg/robin:latest
  71. > ```
  72. > Or mount a local directory, if you want the JSON files on your own disk:
  73. > ```bash
  74. > docker run --rm \
  75. > -v "$(pwd)/.env:/app/.env" \
  76. > -v "$(pwd)/investigations:/app/investigations" \
  77. > --add-host=host.docker.internal:host-gateway \
  78. > -p 8501:8501 \
  79. > apurvsg/robin:latest
  80. > ```
  81. > Investigations are saved to the `investigations/` folder in your working directory and can be loaded from the **Past Investigations** panel in the sidebar.
  82. >
  83. > Create the folder yourself first (`mkdir -p investigations`) so Docker does not create it as root. The container runs as UID 1000, and on Linux a folder owned by anyone else is read-only to it β€” Robin prints one warning and the save fails, though the investigation still runs. If you hit that, or your own UID isn't 1000, use the named volume above, or see [Saved investigations fail with permission denied on Linux](TROUBLESHOOTING.md#saved-investigations-fail-with-permission-denied-on-linux).
  84. - Open your browser and navigate to `http://localhost:8501`
  85. ### Build it yourself
  86. To run your own build instead of the published image, clone the repository and
  87. build it:
  88. ```bash
  89. docker build -t robin .
  90. ```
  91. Then use the same run commands with `robin` in place of `apurvsg/robin:latest`:
  92. ```bash
  93. docker run --rm \
  94. -v "$(pwd)/.env:/app/.env" \
  95. --add-host=host.docker.internal:host-gateway \
  96. -p 8501:8501 \
  97. robin
  98. ```
  99. ```bash
  100. docker run --rm \
  101. -v "$(pwd)/.env:/app/.env" \
  102. -v "$(pwd)/investigations:/app/investigations" \
  103. --add-host=host.docker.internal:host-gateway \
  104. -p 8501:8501 \
  105. robin
  106. ```
  107. - Open your browser and navigate to `http://localhost:8501`
  108. ---
  109. ## Robin as MCP
  110. Any LLM service that speaks MCP can run it, using its own model. The sections
  111. below are worked examples for the common hosts; a host that is not listed works
  112. the same way, with whatever wording it uses for "add an MCP server".
  113. ### Claude Code
  114. ```bash
  115. claude mcp add robin -- docker run -i --rm -v robin-investigations:/app/investigations apurvsg/robin mcp
  116. ```
  117. Or check it into the project in `.mcp.json`. A Tor investigation takes minutes,
  118. so raise the per-server timeout:
  119. ```json
  120. {
  121. "mcpServers": {
  122. "robin": {
  123. "command": "docker",
  124. "args": ["run", "-i", "--rm", "-v", "robin-investigations:/app/investigations", "apurvsg/robin", "mcp"],
  125. "timeout": 600000
  126. }
  127. }
  128. }
  129. ```
  130. ### Codex
  131. ```bash
  132. codex mcp add robin -- docker run -i --rm -v robin-investigations:/app/investigations apurvsg/robin mcp
  133. ```
  134. The ChatGPT desktop app shares this host: adding Robin under **Settings β†’ MCP
  135. servers** there writes the same configuration, and its tools appear in Codex
  136. sessions rather than in an ordinary ChatGPT chat.
  137. A Tor search usually takes a couple of minutes, which can outlast Codex's default
  138. MCP tool timeout, so raise the limits in `~/.codex/config.toml`:
  139. ```toml
  140. [mcp_servers.robin]
  141. command = "docker"
  142. args = ["run", "-i", "--rm", "-v", "robin-investigations:/app/investigations", "apurvsg/robin", "mcp"]
  143. tool_timeout_sec = 600
  144. startup_timeout_sec = 60
  145. # Only for non-interactive runs (`codex exec`): Codex declines MCP tool calls
  146. # that would need approval, so pre-approve Robin's read-only tools.
  147. default_tools_approval_mode = "approve"
  148. ```
  149. Pull the image first (`docker pull apurvsg/robin:latest`) so the first start is
  150. not also a download.
  151. ### Claude Desktop
  152. Add this to `claude_desktop_config.json` and restart the app:
  153. ```json
  154. {
  155. "mcpServers": {
  156. "robin": {
  157. "command": "docker",
  158. "args": ["run", "-i", "--rm", "-v", "robin-investigations:/app/investigations", "apurvsg/robin", "mcp"]
  159. }
  160. }
  161. }
  162. ```
  163. ### Hermes
  164. In `~/.hermes/config.yaml`:
  165. ```yaml
  166. mcp_servers:
  167. robin:
  168. command: docker
  169. args: [run, -i, --rm, -v, "robin-investigations:/app/investigations", apurvsg/robin, mcp]
  170. timeout: 600
  171. ```
  172. ### OpenClaw
  173. ```bash
  174. openclaw mcp add robin --command docker --arg run --arg -i --arg --rm --arg -v --arg robin-investigations:/app/investigations --arg apurvsg/robin --arg mcp
  175. ```
  176. Or in the config, under `mcp.servers`:
  177. ```json
  178. {
  179. "mcp": {
  180. "servers": {
  181. "robin": {
  182. "transport": "stdio",
  183. "command": "docker",
  184. "args": ["run", "-i", "--rm", "-v", "robin-investigations:/app/investigations", "apurvsg/robin", "mcp"],
  185. "requestTimeoutMs": 600000
  186. }
  187. }
  188. }
  189. }
  190. ```
  191. ### ChatGPT
  192. The desktop app's **Settings β†’ MCP servers** configures the Codex host that the
  193. app, the Codex CLI and the IDE extension share, so add Robin there and it
  194. appears in Codex sessions: see [Codex](#codex). An ordinary ChatGPT chat cannot
  195. run it, because chat reaches MCP servers through connectors that run in OpenAI's
  196. infrastructure rather than on your machine.
  197. ### Security
  198. - Nothing sends, executes, or runs a shell. The only writes are saved reports,
  199. into `investigations/` under a filename Robin picks; the agent never names a
  200. path.
  201. - Searches, page scrapes and search-engine health checks go through Tor, so the
  202. app you are using never fetches a dark web page itself. Model provider calls
  203. and the model list refresh go directly, not through Tor.
  204. - Scraped text comes back inside untrusted-data delimiters, with control,
  205. zero-width and bidi characters stripped first, so your model reads it as
  206. evidence rather than as instructions.
  207. ---
  208. ## Acknowledgements
  209. - Idea inspiration from [Thomas Roccia](https://x.com/fr0gger_) and his demo of [Perplexity of the Dark Web](https://x.com/fr0gger_/status/1908051083068645558).
  210. - Tools inspiration from my [OSINT Tools for the Dark Web](https://github.com/apurvsinghgautam/dark-web-osint-tools) repository.
  211. - LLM Prompt inspiration from [OSINT-Assistant](https://github.com/AXRoux/OSINT-Assistant) repository.
  212. - Logo Design by my friend [Tanishq Rupaal](https://github.com/Tanq16/)

Ratings and comments

Sign in to rate this skill.

Sign in

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