AI & ML

yeachan-heo Skill

Nuclear-grade 16-agent pre-publish release gate. Runs /get-unpublished-changes to detect all changes since last npm release, spawns up to 10 ultrabrain agents for deep per-change analysis, invokes /review-work (5 agents) for holistic review, and 1 oracle for overall release synthesis. Use before EVERY npm publish. Triggers: 'pre-publish review', 'review before publish', 'release review', 'pre-release review', 'ready to publish?', 'can I publish?', 'pre-publish', 'safe to publish', 'publishing review', 'pre-publish check'.

ymeiri Bundle

Draft or rewrite user-facing text using the voice-layer model: calibrated personal voice, channel shape, audience adaptation, requested vibe, documentation style, and AI-tell cleanup. Use for Slack, chat, email, PR descriptions, review comments, issues, Confluence, Google Docs, design docs, RFCs, ADRs, release notes, and docs when the user says write in my voice, make this sound like me, rewrite this, polish this, draft a reply, make it corporate-friendly, use a specific vibe, adapt for an audience, de-AI this, or similar. Do not use for code, command output, structured data, verbatim quotes, or voice calibration; use calibrate-my-voice for setup.

ymeiri Bundle

Build or refresh the local voice-layer profile used by write-in-my-voice from user-approved evidence. Use only when the user explicitly asks to calibrate, train, set up, refresh, rebuild, or update their voice-layer profile from sources like git commits, PRs, code reviews, Slack, chat exports, sent email, Google Docs, Confluence, design docs, RFCs, ADRs, issue comments, agent session transcripts, or pasted samples. Do not invoke for ordinary rewrites or drafts.

zaks-io Bundle

Use for workflow setup when setting up or refreshing a repository for agent workflows by creating docs/agents/workflow/config.md with repo commands, planning artifacts, issue tracking, agent adapters, review gates, and environment safety rules.

zasfe Bundle

Socratic method teaching skill that guides users to discover answers themselves through questioning, never giving direct answers. TRIGGER when: user's message contains 'socratic', 'Socrates', or '소크라테스'. Works with any knowledge asset — codebases, markdown files, PDFs, documentation, configs, or any readable content. Respond in the user's language.

holobiomicslab Skill

Use when you have loaded a raw GCIMS dataset and need to isolate the region of interest in drift time (typically 5–16 ms for small organic molecules) to exclude low-drift-time chemical noise, high-drift-time tail artifacts, or off-scale ion signals that would degrade subsequent alignment and peak.

holobiomicslab Skill

Use when integrating with an external API (such as TensorFlow Serving) where changes to the response schema could break dependent code, or when model metadata must be extracted and verified before being used in downstream analysis steps.

holobiomicslab Skill

Use when after calculating mean intensity values across all spectra in an MSI dataset (or within a manually selected ROI), and you need to store the resulting m/z–intensity table in a portable format for downstream ROI analysis, database annotation, or external statistical pipelines.

holobiomicslab Skill

Use when you have a set of compounds (as SMILES strings or molecular structures) that need retention order predictions in a reversed-phase liquid chromatography (RPLC) system at eluent pH ~2.7, and you want to quantify prediction uncertainty rather than relying on a single model's output.

holobiomicslab Skill

Use when after sending HTTP requests to API endpoints (such as /classify or /model/metadata on an NP-Classifier server) to verify that the response is parseable JSON and contains the expected output fields and metadata before attempting to extract or process the data programmatically.

holobiomicslab Skill

Use when you have generated consensus metabolic reconstructions for multiple members of a microbial community (e.

holobiomicslab Skill

Use when when you have loaded a dataset of molecular fingerprint vectors (such as biosynfoni fingerprints from a Zenodo deposit) and need to quantify how sparse the bit-representations are—that is, what fraction of bit positions are zero across the fingerprint collection.

holobiomicslab Skill

Use when you have electron ionization mass spectrum data (m/z and intensity pairs) and a pre-trained transformer model checkpoint, and you need to predict molecular weight directly from the spectrum without manual feature engineering or rule-based methods.

holobiomicslab Skill

Use when when you have a large list of SMILES strings but lack corresponding experimental mass spectra, and you need to generate 30,000+ training pairs to train a deep learning model for molecular structure prediction.

holobiomicslab Skill

Use when you have raw untargeted metabolomics data (at least 3 samples in mzML, mzXML, or CDF format) from qTOF, orbitrap, or Fourier transform ion cyclotron resonance mass analyzers and need to obtain optimized XCMS processing parameters tailored to your specific instrument and dataset rather than.

holobiomicslab Skill

Use when you have preprocessed mass spectrometry fragmentation data (neutral losses and fragment masses extracted and noise-filtered) and want to discover hidden structural motifs across a spectral dataset in an unsupervised manner.

holobiomicslab Skill

Use when you have split multi-assay LC-MS intensity data into training (90%) and test (10%) subsets with assay-specific column prefixes, and you need to fit a discriminant or regression model that respects the block structure (separate assays) while jointly predicting a phenotypic outcome (e.

holobiomicslab Skill

Use when after a TCN-based formula prediction model has generated initial formula candidates from MS/MS spectra, apply this skill to rescore and refine those candidates when you need to improve ranking accuracy.

holobiomicslab Skill

Use when when you have trained a candidate model (e.g., an ensemble, a new architecture) and need to demonstrate its advantage over published or reference implementations on the same test data.

holobiomicslab Skill

Use when when you have an unknown MS/MS spectrum (m/z and intensity pairs) and need to assign a chemical formula and ionization adduct to the precursor mass, particularly when spectrum database lookups are unavailable or when you want to exploit learned patterns in fragmentation rather than.

holobiomicslab Skill

Use when after training a DeepMSProfiler model and generating per-sample predictions: when you need to display Pearson or Spearman correlation coefficients between individual metabolite signals and disease class labels in a matrix form suitable for publication or exploratory review of.

holobiomicslab Skill

Use when you have deployed a microservice (e.g., TensorFlow Serving, REST API) and need to verify that specific endpoints (e.g., /model/metadata, /classify) return responses with correct schema, field names, and data types before consuming them in production workflows or downstream applications.

holobiomicslab Skill

Use when when analyzing LC-MS data from stable isotope labeling experiments where measured isotopologue abundances are contaminated by naturally occurring isotopes and tracer isotopic impurity, and you have access to unlabeled sample reference measurements to empirically model these confounding.

holobiomicslab Skill

Use when you have two LC-MS feature tables (each with m/z, retention time, and intensity columns) from separate metabolomic experiments or replicates, and you need to establish which features in dataset A correspond to which features in dataset B to enable comparative or longitudinal analysis.

holobiomicslab Skill

Use when you have a metabolomics dataset and want to perform pathway enrichment analysis using ORA, but need to first understand its behavior, limitations, and correct application through reproducible simulation.

holobiomicslab Skill

Use when training a Transformer or neural network model on a large dataset (e.g., 80,000+ training samples) where validation performance is monitored to prevent overfitting, and you need to halt training early or recover the -performing model checkpoint without re-executing the entire training loop.

holobiomicslab Skill

Use when when you have an aligned GCIMS dataset and need to configure the findPeaks function with CWT algorithm to detect peaks across retention time and drift time dimensions.

holobiomicslab Skill

Use when you have preprocessed MS/MS spectral data (m/z and intensity arrays) and need to map it to a fixed-size latent vector for use in an encoder–decoder architecture. Specifically applicable when the downstream task requires a molecular structure reconstruction (e.

holobiomicslab Skill

Use when when you have raw or partially processed metabolomics data (feature tables with sample metadata) and need to apply quality-control metrics, normalization, statistical inference, or advanced classification/variable selection without relying on a Galaxy instance.

holobiomicslab Skill

Use when you have raw or preprocessed mass spectrometry feature matrices (e.g., from low mass resolution or sparse acquisition) and want to enhance signal quality and spatial resolution across tissue or single-cell samples.

holobiomicslab Skill

Use when after training a tandem mass spectrometry embedding model (such as MSBERT) on a reference spectral library (e.g., GNPS), apply this skill to confirm that the high-dimensional embedding vectors project into interpretable chemical space.

holobiomicslab Skill

Use when you have preprocessed 1D NMR spectra (¹H and/or ¹³C) and need to extract spectral features for molecular structure inference on molecules with up to 19 heavy atoms. The skill is necessary as the first stage before fragment assembly or connectivity prediction;

holobiomicslab Skill

Use when you have a trained graph neural network model for CCS prediction and need to identify which molecular structural features drive individual predictions or systematic biases.

holobiomicslab Skill

Use when you have a labeled dataset of DIA raw files (.raw, .d, .wiff) with known quality annotations and have extracted the 15 iDIA-QC metrics (raw file characteristics from timsTOF, TripleTOF, or Orbitrap instruments).

holobiomicslab Skill

Use when when you have quantified intracellular metabolite abundances (LC-MS normalized values) from multiple samples and need to predict how differences in substrate availability translate into differences in metabolic flux for specific reactions in a constraint-based metabolic model.

holobiomicslab Skill

Use when you have annotated LC-MS ROI snippets with ground-truth peak/non-peak labels and boundary coordinates (peak start/end positions), and you need to build a model that can discriminate true peaks from false peaks while precisely localizing peak boundaries for area integration in future LC-MS.