花 · 桜 · 春
proto.type · research mode

Peer-reviewed
before you wake up.

General AI tools don't understand peer review standards, statistical validity, or what makes a result reproducible. Research Mode gives you a team of specialists who do — overnight, on your data, in your domain.

proto.type · Research Morning Report · overnight run complete authored by Noa · Seren · Petra
the question
research_intent
“Does the L452R spike mutation in this isolate confer increased ACE2 binding affinity compared to the reference strain?”
domain · virology    confidence target · p < 0.05    replicates · 3
Noa · Domain Scientist
Seren · Statistician
Petra · Peer Review
Cato · Lab Systems
Finn · Data Integrity
overnight
23:14
Lumen
41 papers reviewed. 3 prior studies on L452R ACE2 binding. Consensus: moderate affinity increase expected.
01:38
Cato
Binding affinity pipeline complete. Docking simulations across 3 replicates. Raw scores logged and verified.
03:52
Seren
ANOVA complete. F(2,6) = 18.4, p = 0.003. Effect size Cohen’s d = 1.61. Power: 0.94 at α = 0.05.
05:21
Petra
Methodology survives peer review. One additional replicate recommended before submission. Finding is real.
morning report
reproducible — approaching publication-ready
L452R significantly increases ACE2 binding affinity.
p-value0.003
effect sized = 1.61
replicates3 / 3
power0.94
Petra · peer review assessment
“"Controls are present and appropriate. Statistical treatment is correct. One additional replicate would strengthen the claim for submission. The finding stands."”
research integrity 96 all data traced · all steps documented
research_intent
The constitutional field
for your science.
Before a single analysis runs, Noa interrogates your intent. What is the question? What would a valid result look like? What would invalidate it? What is the peer review standard? Every agent is bound to these answers the same way the build team is bound to the Mission Object.
A vague intent produces vague findings. A precise intent produces findings you can defend.
question
"Does Compound X inhibit protein Y expression in HeLa cells at concentrations below cytotoxic threshold?"
valid result
Statistically significant inhibition (p < 0.05) across at least 3 independent replicates with consistent dose-response curve.
invalidating conditions
>20% cell viability loss at effective concentration. Any single-replicate finding presented as conclusive.
peer review standard
Methods section reproducible by an independent lab. All raw data preserved and citable.
ethics constraints
All compounds under BSL-2 protocol. Ren flags any deviation before analysis begins.
the research roster

Seven specialists.
Every discipline.

Domain science, statistics, lab systems, peer review, ethics, data integrity, literature synthesis. The complete analytical team — overnight, on your question.

Hebrew
Noa
motion, flow
Domain Scientist
Validates findings against established science. Knows the difference between surprising and impossible. Required for research planning.
Welsh
Seren
star
Statistician
Expert in experimental design and statistics. Never claims significance without proper testing. Every result needs her approval.
Latin
Cato
knowing
Lab Systems Engineer
Builds the code that runs experiments — pipelines, instrument connections, analysis. Understands both science and engineering.
Greek
Petra
rock
Peer Review Critic
Reviews work like the harshest journal reviewer. Catches confounds, missing controls, overclaimed conclusions. No finding ships without her approval.
Japanese · 蓮
Ren
lotus
Research Ethics Officer
Expert in research ethics and regulation — IRB, IACUC, safety levels. Knows the specific rules for your domain. Her concerns are blocking.
Irish
Finn
fair, bright
Data Integrity Officer
Obsessed with reproducibility. Traces every data point to its source. Documents every transformation. Creates the reproducibility report.
Latin
Lumen
light
Literature Synthesiser
Surveys existing research at scale. Finds consensus, open debates, and gaps. Every claim is situated in the literature.
Activates when your
project is research.
Any domain.
Any instrument.
supported domains

Your field.
Your team.

Research Mode activates on intent, not a list of approved domains. If you are asking a falsifiable question with data, the team can help. The MCP server connects to your instruments and databases.

Molecular Biology
genomics · proteomics · sequencing
Cato builds the analysis pipeline. Noa validates against the literature. Seren handles the statistics. Overnight from raw sequencing output to a morning report you can defend.
Drug Discovery
chemistry · pharmacology · safety
Compound safety profiles, dose-response analysis, regulatory constraints. Ren flags BSL and REACH requirements before a line of analysis runs. Petra reviews before anything is reported.
Clinical Research
epidemiology · trials · biostatistics
Seren designs the statistical model. Petra reviews the methodology. Finn traces provenance. IRB-aware from day one. Ren is required on any project with human subjects.
Any Field
materials · climate · social science · physics
Research Mode activates on project intent. If you are asking a falsifiable question with data, the team adapts. Noa will read whatever literature exists in your domain.
← Back to main page
created by asha.software · intentional software · built with love and respect for user privacy
花 · 桜 · 春
language