PROMPTR™
Now included in all DECODR Premium Accounts - click on the figure below for a quick tour!
PROMPTR is DECODR’s new AI‑ready export - a complete, self‑describing download of your entire analysis history, built so tools like Claude, ChatGPT, or Copilot can actually reason over your data. Not a raw dump, not a pile of spreadsheets - a structured, annotated, schema‑defined export that explains itself and cites the chromatograms behind every answer.
It turns months or years of CRISPR work into something you can interrogate with a single question.
What PROMPTR actually gives you
PROMPTR exports every analysis in your DECODR account into one file:
-
Flat, joinable tables for every part of the assay
-
A data dictionary defining every column
-
A readme explaining the assay, caveats, and common traps
-
Outputs link back to DECODR so every AI generated insight is traceable to the underlying chromatograms
-
Built in guidance (e.g., “filter paired-end runs before aggregating,” “treat R² < 0.6 with caution,” “blank ≠ zero”)
It’s the first export designed for AI: meaning the model doesn’t need a human to explain the schema before it can answer real scientific questions.
Questions that analyze the full account in one unified query:
-
"Rank all samples by knockout score and provide a top 10 leaderboard.”
-
"Which samples contain unusually large deletions, over 100 bp?"
-
"Compare knock-in efficiency across all HDR experiments."
Prompts for visualization and deep dives:
-
"Build a dashboard summarizing all editing outcomes."
-
"What are the most important trends in my CRISPR results?"
-
"Compare every guide I've used for NRF2 and tell me which is most effective at introducing frame-shift indels."
-
“Review my experiments as if you were a journal reviewer. What concerns would you raise and how could they be addressed?”
These are one‑line prompts - and they work because PROMPTR provides the structure an AI needs.
Researchers
Ask your entire history a question. Instead of opening dozens of result pages one at a time, you can now ask:
-
“Which guide gave me the highest knockout score across all experiments?”
-
“How often did my dual‑guide runs drop out?”
-
“Which cell lines show the strongest frameshift bias?”
PIs and facility directors
Your data leaves with you. Open format, no lock‑in, instantly usable in:
-
Lab notebooks
-
Manuscripts
-
Grant reports
-
Internal QC summaries
Technical buyers
Most exports are unusable by LLMs until a human explains them. PROMPTR ships with its own schema and caveats, so the model aggregates correctly instead of double‑counting paired-end reads or misinterpreting blanks.