geo-sleuth
geo-sleuth is an open-source agent skill from Oldcircle that helps coding agents work out where a photo was taken, using map, elevation and imagery data.
No visits from OpenBeat in the last 30 days yet. Every click on “Visit” is counted here.
Pricing and API checked by an OpenBeat editor on Nov 10, 2026 · facts from the maker’s own site ·
Tool insights
geo-sleuth gives a coding agent a set of scripts for locating a photo. It reports a camera position, heading and error radius, and it produces a satellite evidence image.
Working out where a photo was taken by hand means switching between map data, terrain data, satellite views and street-level imagery. geo-sleuth packages those steps as scripts that an agent can run and combine.
Worth a look for developers and researchers who already use a coding agent and want help with photo geolocation. Not a fit if you don't use an agent that reads SKILL.md and runs shell commands, or if you need a guaranteed, repeatable result.
What it does
geo-sleuth is an agent skill made by Oldcircle. It helps an AI coding agent find the place where a photo was taken. The agent works with OpenStreetMap geometry, elevation data, satellite imagery and street view.
The output is an estimate of the camera position, the heading and an error radius. It also produces a satellite evidence image. The facts do not say what format that image takes or how it is laid out.
Key features
- Finds a photo's location by matching OpenStreetMap geometry.
- Uses elevation data for skyline matching.
- Searches with satellite imagery and street view.
- Reports camera position, heading and error radius.
- Produces a satellite evidence image.
- Ships about twenty single-purpose Python scripts.
- Follows the standard Agent Skill folder layout with a SKILL.md file.
- Installs with the
npx skills addcommand.
The listed models are CLIP and DINOv2. The facts do not say exactly how each one is used.
Pricing
geo-sleuth is open source under the MIT license. As of November 2026, no paid plans are listed. Whether a free plan exists in the usual sense isn't stated on the maker's site. Costs of the agent you run it with, or of any data sources, are not covered in the facts.
API and integrations
No hosted API is stated on the maker's site, and no SDKs are listed. The skill runs as a command-line toolset. It works with agents that can read SKILL.md and run shell commands. Listed integrations are Claude Code, Codex, Cursor, Gemini CLI, OpenCode and GitHub Copilot. The only listed platform is CLI.
Good to know
- Results depend on the model and the decisions it makes during a case.
- A different model may not reach the same result.
- You need an agent that reads SKILL.md and runs shell commands.
- No documentation page, pricing page or changelog is linked. The GitHub repository is the only listed source.
- The founding year isn't stated on the maker's site.
FAQ
What is geo-sleuth?
geo-sleuth is an open-source agent skill from Oldcircle. It helps a coding agent find where a photo was taken, using OpenStreetMap, elevation data, satellite imagery and street view.
Is geo-sleuth free and open source?
Yes, it is open source under the MIT license. As of November 2026, no paid plans are listed. Any costs from the agent or model you run it with are not covered in the facts.
Which agents does geo-sleuth work with?
The listed integrations are Claude Code, Codex, Cursor, Gemini CLI, OpenCode and GitHub Copilot. Any agent that reads SKILL.md and runs shell commands is needed.
How do I install geo-sleuth?
It installs with the `npx skills add` command. It is a standard Agent Skill folder with a SKILL.md file and about twenty Python scripts.
Will geo-sleuth always find the same location?
Not necessarily. Results depend on the model and its decisions, and a different model may not reach the same result.
Releases
No tracked releases yet.
In the news
No coverage yet.