AI agent for agricultural scientists
Herbicide Resistance Screening Tracker Agent
Make every resistance call rest on a valid assay with consistent data
What it does
A resistance screen exposes weed populations to a dose series and counts survivors. Notes are in different formats and a failed control can be missed, so a false resistance call reaches a report. This agent tracks every population. It logs each weed population with its source field, doses and survival counts, checks that the controls behaved as expected, flags failed assays and queues repeat runs. It compares the repeat results with the first and rechecks the counts. It also calculates survival at each dose and compares with the sensitive standard. The scientist approves every resistance call. Edge case: the susceptible control survived. The agent marks the whole assay as invalid.
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.
Read the steps as a list
- New assay results entered
- Load population data, doses and survival counts
- Check the data format and fill missing fields with queries to the technician
- Did the untreated and susceptible controls behave as expected?If not: Mark the assay invalid and queue a repeat. Back to step 2.
- Calculate survival at each dose and compare with the susceptible standard
- Classify each population as susceptible, suspected or resistant
- Is a suspected population confirmed in a second valid run?If not: Queue a repeat run and hold the call. Back to step 3.
- Scientist approves each resistance callThe agent waits here for your OK.
- Update the resistance register
- Resistance register and repeat queue
How it decides
An assay is valid only if the untreated control grows well and the susceptible standard dies at the expected dose. A resistance call needs survival above the cut-off in two valid runs.
- Untreated control survival must be over 90 percent
- The susceptible standard must show under 10 percent survival at the discriminating dose
- A resistance call needs two valid runs
- Survival above 20 percent at the discriminating dose is suspected resistance
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Control cut-offs
- Discriminating doses
- Survival threshold (default 20 percent)
- Number of runs required (default 2)
- Data sheet format
What keeps you in control
It always asks you first
- Scientist approves every resistance call before it is recorded or reported
Hard limits
- Never reports resistance from a single run
- Keeps raw counts with every call
It stops when
- Done: all populations classified from valid runs
- Stop: seed or plant supply insufficient for repeats, ask the scientist
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.
- 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
