Installable skillMIT

K-Dense-AI Scientific Computing Bundle

- `alphafold-database` - Protein structure prediction

BY plurigridUPDATED Oct 27, 2026
INTENDED FOR>_CodexInstall in a listed agent.

instructions.md

---
name: k-dense-ai
description: '- `alphafold-database` - Protein structure prediction'
---
# K-Dense-AI Scientific Computing Bundle

> Provenance: K-Dense-AI / Codex scientific skills
> GF(3) Trit: **+1 (PLUS)** - Generative/Synthetic
> Mutual Awareness: `trailofbits-security` bundle (MINUS -1)

## Skills (37)

### Bioinformatics Core
- `alphafold-database` - Protein structure prediction
- `biopython` - Biological computation toolkit
- `esm` - Evolutionary Scale Modeling
- `scanpy` - Single-cell analysis
- `anndata` - Annotated data matrices
- `cellxgene-census` - Cell atlas queries

### Databases
- `chembl-database` - Bioactive molecules
- `pubchem-database` - Chemical compounds
- `uniprot-database` - Protein sequences
- `pdb-database` - Protein structures
- `kegg-database` - Pathway maps
- `reactome-database` - Biological pathways
- `brenda-database` - Enzyme data
- `drugbank-database` - Drug information
- `gene-database` - Gene annotations
- `biorxiv-database` - Preprints
- `clinicaltrials-database` - Clinical studies

### Chemistry & Drug Discovery
- `rdkit` - Cheminformatics
- `deepchem` - Deep learning for chemistry
- `catalyst-chemical` - Reaction modeling

### Molecular Simulation
- `enzyme-autodiff` - Automatic differentiation for enzymes

### Lab & Workflow
- `benchling-integration` - Lab notebook
- `bioservices` - Web service APIs
- `adaptyv` - Adaptive experiments
- `biomni` - Multi-omics

### Physics & Quantum
- `astropy` - Astronomy
- `cirq` - Quantum circuits

### ML for Science
- `scikit-bio` - Biological data science
- `scikit-learn` - Machine learning
- `scikit-survival` - Survival analysis
- `arboreto` - Gene regulatory networks
- `aeon` - Time series for science

### Categorical
- `phylogenetic-operad-acset` - Tree structures as operads

## Mutual Awareness Protocol

```clojure
{:bundle "k-dense-ai"
 :trit :plus
 :aware-of ["trailofbits-security"]
 :interface
 {:analyze (fn [molecule] "Use rdkit/deepchem")
  :validate (fn [result] "Delegate to trailofbits for security review")
  :synthesize (fn [target] "Generate candidates")}
 :handoff-to "trailofbits-security"
 :handoff-trigger [:untrusted-input :external-api :file-parsing]}
```

## Usage

```bash
# Load bundle
skill k-dense-ai

# Cross-bundle workflow
skill k-dense-ai -> trailofbits-security  # Validate drug candidate data pipeline
```
COMPATIBILITY & REFERENCES

Platform guidance

Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.