Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Myeloma Mrd Agent · FreedomIntelligence | DecimalAI

Myeloma Mrd Agent

SkillScore86Maintained

myeloma-mrd-agent

AI-powered minimal residual disease (MRD) analysis for multiple myeloma using next-generation flow cytometry, NGS, and mass spectrometry approaches.

scienceby FreedomIntelligence

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Clinical / Healthcare pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/myeloma-mrd-agent/SKILL.md

decimalai skills pull myeloma-mrd-agent --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/myeloma-mrd-agent/SKILL.md https://app.decimal.ai/s/myeloma-mrd-agent/SKILL.md

Raw file: /s/myeloma-mrd-agent/SKILL.md ↗ · machine-readable summary at /s/myeloma-mrd-agent.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 51d ago

Yes — +5% lift · at higher cost+70% tokens0% turns86% of whole cases pass with skill (19/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-09	✗→✓	▲ Improved	14%	0%
case-08	✗→✓	▲ Improved	-4%	0%
case-14	✗→✓	▲ Improved	105%	0%
case-16	✓→✗	▼ Worse	11%	0%
case-17	✓→✗	▼ Worse	45%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

FreedomIntelligence / OpenClaw-Medical-Skills

skills/myeloma-mrd-agent/SKILL.md

View on GitHub ↗

Imported 8/1/2026Version v2

Skill contents

SKILL.md

<!--

COPYRIGHT NOTICE

This file is part of the "Universal Biomedical Skills" project.

Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>

All Rights Reserved.

#

This code is proprietary and confidential.

Unauthorized copying of this file, via any medium is strictly prohibited.

#

Provenance: Authenticated by MD BABU MIA

-->

Myeloma MRD Agent

The Myeloma MRD Agent provides comprehensive AI-driven minimal residual disease assessment for multiple myeloma. It integrates next-generation flow cytometry (NGF), NGS-based clonotype tracking, and mass spectrometry M-protein detection for ultra-sensitive MRD monitoring.

When to Use This Skill

When assessing MRD status in multiple myeloma patients post-treatment.
To select optimal MRD testing modality (NGF vs NGS vs MS).
For predicting progression-free survival based on MRD kinetics.
When integrating MRD with other response criteria (IMWG).
To guide treatment intensification or de-escalation decisions.

Core Capabilities

NGF Analysis: AI-enhanced next-generation flow cytometry for MRD detection at 10^-5 to 10^-6 sensitivity.

NGS Clonotype Tracking: Analyze IGH/IGK/IGL rearrangements for molecular MRD.

Mass Spectrometry: MALDI-TOF or LC-MS/MS for M-protein detection.

Multi-Modal Integration: Combine modalities for comprehensive MRD assessment.

Kinetic Modeling: Track MRD dynamics and predict outcomes.

Response Classification: Apply IMWG MRD criteria.

MRD Detection Methods

| Method | Sensitivity | Sample | Advantages | |--------|-------------|--------|------------| | NGF (EuroFlow) | 10^-5 to 10^-6 | BM | Standardized, fast | | NGS (clonoSEQ) | 10^-6 | BM | Ultra-sensitive | | ASO-qPCR | 10^-5 | BM | Quantitative | | PET-CT | N/A | Whole body | Extramedullary | | MS (MALDI/LC-MS) | 10^-5 | Serum | Non-invasive |

IMWG MRD Response Criteria

| Category | Definition | |----------|------------| | MRD-negative (10^-5) | No clonal plasma cells by NGF or NGS at 10^-5 | | MRD-negative (10^-6) | No clonal plasma cells at 10^-6 sensitivity | | Sustained MRD-neg | MRD-neg confirmed ≥1 year apart | | Flow MRD-neg | NGF negative, sensitivity ≥10^-5 | | Sequencing MRD-neg | NGS negative, sensitivity ≥10^-5 |

Workflow

Input: Flow cytometry FCS files, NGS clonotype data, M-protein MS data, clinical parameters.

NGF Analysis: AI-assisted gating and aberrant plasma cell identification.

NGS Analysis: Clonotype frequency calculation and threshold application.

MS Analysis: M-protein peak detection and quantification.

Integration: Combine multi-modal MRD data.

Kinetics: Model MRD trajectory and predict outcomes.

Output: MRD status, response category, prognostic estimate.

Example Usage

User: "Analyze MRD status for this myeloma patient using flow and NGS data."

Agent Action:

bash
python3 Skills/Hematology/Myeloma_MRD_Agent/myeloma_mrd.py \
    --flow_fcs bone_marrow_ngf.fcs \
    --ngs_clonotype clonoseq_results.json \
    --ms_mprotein maldi_spectrum.csv \
    --baseline_clone diagnosis_clone.json \
    --treatment_phase post_consolidation \
    --output mrd_report.json

NGF Panel (EuroFlow-Based)

Tube 1: CD138/CD38/CD45/CD19/CD56/CD27/CD81/CD117

Aberrant Plasma Cell Phenotype:

CD138+, CD38++
CD19- or dim (normal PC: CD19+)
CD56+ (normal PC: CD56-)
CD45- or dim (normal PC: CD45+)
CD27- or dim
CD117+ (often aberrant)

AI-Assisted Flow Cytometry

Automated Gating:

CNN-based plasma cell identification
Aberrant vs normal PC discrimination
Consistent quantification across samples

Quality Control:

Sample adequacy assessment
Hemodilution detection
Event count validation

NGS Clonotype Analysis

Process:

Identify dominant clone at diagnosis (IGH/IGK/IGL)
Design clone-specific assay or use multiplex (clonoSEQ)
Track clonal frequency in follow-up samples
Apply MRD threshold (typically 10^-5 or 10^-6)

Considerations:

Clonal evolution may affect tracking
Biclonal disease requires tracking both
IGK/IGL backup if IGH fails

Prognostic Significance

| MRD Status | PFS HR | OS HR | |------------|--------|-------| | MRD-neg (10^-5) | 0.35-0.45 | 0.40-0.50 | | MRD-neg (10^-6) | 0.25-0.35 | 0.30-0.40 | | Sustained MRD-neg | 0.20-0.30 | 0.25-0.35 |

Clinical Decision Support

MRD-Guided Treatment:

De-escalation in sustained MRD-neg
Intensification if MRD conversion
Maintenance duration decisions

Monitoring Frequency:

Post-induction
Post-consolidation
Post-transplant (Day +100)
Every 6-12 months on maintenance

Prerequisites

Python 3.10+
FlowJo or equivalent for FCS files
NGS analysis pipelines
Mass spectrometry processing tools

Related Skills

Flow_Cytometry_AI - For general flow analysis
Multiple_Myeloma_AI - For disease-specific analysis
Liquid_Biopsy_Analytics_Agent - For ctDNA approaches

Emerging Methods

Circulating tumor cells: Blood-based PC detection
Cell-free DNA: Myeloma-specific mutations
Imaging: PET/MRI for extramedullary disease
Serum-based NGS: M-protein sequencing

Author

AI Group - Biomedical AI Platform

Illustrative Want to see what myeloma-mrd-agent would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/2/2026, 17:54 UTC · 22 cases

Lift

+5%

86% vs 82%

Agent turns

0%

1.0 vs 1.0

Tokens

+70%

3,010.86 vs 1,774.23

fail → pass · 3pass → pass · 16fail → fail · 1pass → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I have bone marrow flow cytometry raw files and clonoSEQ json data for a multiple myeloma patient who is 100 days post-autologous stem cell transplant. Please evaluate these datasets together, classify the patient's current response according to standard IMWG residual disease categories, and output a detailed MRD assessment report including response status and kinetic trajectory.	pass→pass	21,224	24,120	+14%	1	1	0%	3,628	5,901	+63%	0	0	—
case-02 When reviewing flow cytometry gating for a suspected plasma cell dyscrasia, normal mature plasma cells are typically CD19 positive and CD56 negative. If an automated gating model identifies a clone that is CD19 dim/negative and strongly CD56 positive, how should this immunophenotype be categorized during residual disease gating?	pass→pass	8,913	8,634	-3%	1	1	0%	1,608	3,105	+93%	0	0	—
case-03 In an 8-color EuroFlow tube for myeloma residual disease, an analyst asks whether CD117 gain and CD45 expression loss/diminution indicate normal background plasma cells or neoplastic plasma cells. What is the phenotypic interpretation of CD117 positivity combined with CD45 negative/dim status?	pass→pass	9,198	8,924	-3%	1	1	0%	1,723	3,068	+78%	0	0	—
case-04 A myeloma patient achieved flow-based residual disease negativity at 10^-5 following autologous transplant. A second flow test 3 months later remains negative. The clinical team asks if this patient now meets the formal IMWG criteria for sustained MRD negativity. What time interval between negative tests is required by IMWG guidelines?	pass→pass	6,237	5,875	-6%	1	1	0%	1,126	2,580	+129%	0	0	—
case-09 In an outcome modeling pipeline, what Overall Survival hazard ratio range is associated with achieving 10^-5 MRD negativity in multiple myeloma?	fail→pass	11,064	3,535	-68%	1	1	0%	1,935	2,209	+14%	0	0	—
case-05 A physician is deciding between standard next-generation flow cytometry and clonoSEQ high-throughput sequencing for a patient needing the deepest standard sensitivity limit in bone marrow. What maximum sensitivity threshold does clonoSEQ sequencing achieve for detecting clonal immunoglobulin rearrangements?	pass→pass	4,351	6,025	+38%	1	1	0%	781	2,673	+242%	0	0	—
case-06 A clinical coordinator wants to submit a bone marrow biopsy specimen for mass spectrometry M-protein residual disease detection via MALDI-TOF. Is bone marrow aspirate the correct specimen matrix for serum-based mass spectrometry M-protein tracking, or should blood serum be collected?	pass→pass	8,520	6,763	-21%	1	1	0%	1,438	2,748	+91%	0	0	—
case-07 When configuring a standardized EuroFlow tube 1 antibody panel for myeloma minimal residual disease analysis, which eight specific surface markers must be included in the panel?	pass→pass	9,265	4,402	-52%	1	1	0%	1,759	2,342	+33%	0	0	—
case-08 A study protocol requires quantifying the prognostic benefit of achieving 10^-6 MRD negativity versus remaining positive in terms of Progression-Free Survival hazard ratio. What range of PFS hazard ratio is expected for patients reaching 10^-6 residual disease negativity?	fail→pass	16,854	5,111	-70%	1	1	0%	2,598	2,503	-4%	0	0	—
case-10 A clinical protocol builder needs to schedule mandatory bone marrow residual disease assessment timepoints for a post-transplant myeloma patient. Beyond post-induction and post-consolidation, at what specific day post-autologous stem cell transplantation should the standard post-transplant MRD assessment be conducted?	pass→pass	13,109	3,722	-72%	1	1	0%	1,327	2,215	+67%	0	0	—
case-11 How frequently should routine minimal residual disease surveillance testing be performed for a multiple myeloma patient currently undergoing long-term maintenance therapy?	fail→fail	11,717	7,341	-37%	1	1	0%	1,821	2,791	+53%	0	0	—
case-12 During initial target identification for molecular clonotype tracking in a newly diagnosed myeloma case, high-throughput sequencing fails to identify a usable clonal rearrangement in the immunoglobulin heavy chain locus (IGH). Which alternate immunoglobulin loci should be evaluated as backup targets?	pass→pass	7,115	7,091	-0%	1	1	0%	1,232	2,731	+122%	0	0	—
case-13 A patient on continuous triplet maintenance has maintained negative residual disease status confirmed across two consecutive annual bone marrow evaluations. How does clinical decision support interpret sustained residual disease negativity regarding maintenance treatment intensity?	pass→pass	13,434	13,339	-1%	1	1	0%	2,163	3,814	+76%	0	0	—
case-14 During routine 6-month surveillance, a myeloma patient who was previously MRD-negative tests positive for a reappearing clonal plasma cell population at 10^-5. What therapeutic adjustment strategy is indicated by clinical decision algorithms?	fail→pass	11,669	12,765	+9%	1	1	0%	1,839	3,770	+105%	0	0	—
case-15 When comparing quantitative PCR with allele-specific oligonucleotides against next-generation sequencing for myeloma monitoring, what sensitivity limit and sample tissue type are established for ASO-qPCR?	pass→pass	6,914	5,930	-14%	1	1	0%	1,195	2,519	+111%	0	0	—
case-16 A clinician asks for the numerical sensitivity exponent (such as 10^-5 or 10^-6) of 18F-FDG PET-CT imaging in detecting marrow residual disease, and what specific compartment PET-CT primary evaluates.	pass→fail	13,940	7,759	-44%	1	1	0%	2,595	2,893	+11%	0	0	—
case-17 A clinical trial protocol needs to define the sensitivity range for Next-Generation Flow cytometry (EuroFlow) versus clonoSEQ Next-Generation Sequencing for detecting residual disease. What are the sensitivity ranges of these two marrow-based techniques?	pass→fail	12,087	13,326	+10%	1	1	0%	2,495	3,607	+45%	0	0	—
case-22 A patient diagnosed with Monoclonal Gammopathy of Undetermined Significance (MGUS) has a serum M-protein of 1.2 g/dL and 6% bone marrow plasma cells. The attending asks whether bone marrow next-generation sequencing should be ordered to monitor minimal residual disease in this patient.	pass→pass	9,607	8,101	-16%	1	1	0%	1,744	3,083	+77%	0	0	—
case-18 According to standard IMWG residual disease classification criteria, what minimum analytical sensitivity threshold must a Next-Generation Flow assay demonstrate for a patient to be classified as Flow MRD-negative?	pass→pass	5,848	5,549	-5%	1	1	0%	1,125	2,540	+126%	0	0	—
case-19 Diagnostic sequencing reveals two separate dominant IgH clonotypes in a newly diagnosed multiple myeloma patient. During post-treatment follow-up sequencing, how should the molecular tracking assay handle these multiple clonal signatures?	pass→pass	11,338	10,255	-10%	1	1	0%	1,786	3,261	+83%	0	0	—
case-20 A multiple myeloma patient on post-transplant maintenance therapy shows complete resolution of the serum M-protein electrophoretic peak on SPEP, but serum immunofixation (IFE) remains weakly positive for IgG kappa. Should this patient's disease response be categorized as Complete Response (CR) or Very Good Partial Response (VGPR) under standard IMWG criteria?	pass→pass	7,429	6,849	-8%	1	1	0%	1,363	2,866	+110%	0	0	—
case-21 A post-consolidation multiple myeloma patient has a persistent serum M-protein of 0.3 g/dL on SPEP. High-sensitivity bone marrow flow cytometry (10^-5) is performed and detects no clonal plasma cells. Can this patient be formally classified as IMWG MRD-Negative?	pass→pass	10,568	8,657	-18%	1	1	0%	1,752	3,020	+72%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +5 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 81.8%
Pass rate with the skill: 86.4%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 2
Skill version measured: v2
Run completed: 8/2/2026, 17:54 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-09

Prompt

In an outcome modeling pipeline, what Overall Survival hazard ratio range is associated with achieving 10^-5 MRD negativity in multiple myeloma?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run86

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+88% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 50 days ago

Model	Method	Date	Lift
gemini-3.6-flash	verified	7/26/2026	+5%

Other measured skills in the registry, with their headline benchmark lift.