Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for ecologists

Camera Trap Image Review Agent

A verified detection dataset for every camera check, ready for analysis

Camera Trap Image Review Agent: what goes in, what the agent does and what you get

What it does

Camera traps produce huge numbers of images, and most are empty frames triggered by wind or grass. When a batch of cards is uploaded, this agent first checks each camera's metadata: dates, times and site IDs, because a wrong clock ruins activity analysis. Clock errors are corrected from known timestamps or sent to the field team. It then removes empty images, labels animals with an image model and groups images into independent detection events using your time rule. Labels below the confidence threshold, and every rare or protected species detection, go to a review queue. After review, it tallies detections per site and flags cameras with mostly blocked or empty frames. The ecologist reviews the queue and approves the final dataset. Edge case: a camera whose clock reset to the factory date has times corrected from the card's known timestamps, with a note.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueApprovedNo 1 STARTS WHEN Camera cards uploaded 2 USES A TOOL Check dates, times and site IDs for each camera 3 CHECKS THE RESULT Is the metadata valid for every camera? If not: correct clock errors from known timestamps orask the field team. Back to step 2. 4 USES A TOOL Remove empty images and label species 5 DOES Group images into independent detection events 6 DOES Send low-confidence and rare species to review 7 YOU APPROVE Ecologist reviews queue and approves dataset 8 RESULT Detection dataset and camera status report saved
Read the steps as a list
  1. Camera cards uploaded
  2. Check dates, times and site IDs for each camera
  3. Is the metadata valid for every camera?If not: correct clock errors from known timestamps or ask the field team. Back to step 2.
  4. Remove empty images and label species
  5. Group images into independent detection events
  6. Send low-confidence and rare species to review
  7. Ecologist reviews queue and approves datasetThe agent waits here for your OK.
  8. Detection dataset and camera status report saved

How it decides

It accepts labels above the confidence threshold for common species and sends all low-confidence and rare species images to human review.

  • Send every rare or protected species detection to review
  • Count detections as independent after the set gap
  • Flag cameras with mostly blocked or empty frames

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Confidence threshold (default 0.85)
  • Rare and protected species list
  • Independence gap between events (default 30 minutes)
  • Review queue format

What keeps you in control

It always asks you first

  • Final detection dataset
  • Sharing rare species locations

Hard limits

  • Never deletes original images
  • Never publishes rare species locations

It stops when

  • Done: dataset approved
  • Stop: site metadata missing; ask the field team

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA September check of 24 cameras produced 18,000 images. The metadata check failed on camera 17, whose clock had reset to January 1, 2020. The agent corrected it from the setup photo timestamp. It removed 14,200 empty frames and sent 600 images to review, including two possible lynx. The ecologist confirmed one lynx and approved the dataset. Camera 9 was flagged for vegetation blocking.

More agents for ecologists