What it does
This repo is a collection of 36 skills; it is a guide with example skills
- Connects Instruction files the AI reads, with steps that call an outside service
- Account or key Some steps need a key for DMXAPI_API_KEY, OPENROUTER_API_KEY, CHAT_API_KEY, OpenAI, Anthropic, XAI_API_KEY (per the README)
- Category Coding
Maker’s original description
🏭 AI agent platform with skills for protein engineering, the noob-friendly AI tutorial tool for life science professionals. (ACL Demo2025, arXiv 2026)For makers: add a badge to your README
If you made this, paste this code into your README. The badge links back to this page.
[](https://mcp.popnx.com/en/skills/ai4protein-venusfactory2)Use together with
MCP servers that connect the AI to services for the same work
Fallow
Codebase intelligence for TypeScript and JavaScript, exposed as structured MCP analysis tools.
Desktop Commander
MCP server for terminal commands, file operations, and process management
Similar Skills
Same kind of work, ordered like the Skills list
| This one VenusFactory2 | rea | andrej-karpathy-skills | graphify | MoneyPrinterTurbo | |
|---|---|---|---|---|---|
| Maker | ai4protein | morluto | multica-ai | Graphify-Labs | harry0703 |
| Skills | 36 | 1 | 1 | 1 | 1 |
| GitHub stars | 253 | 467,235 | 234,125 | 149,873 | 144,332 |
| Last code change | Nov 18, 2026 | Nov 18, 2026 | Apr 20, 2026 | Nov 18, 2026 | Nov 18, 2026 |
| License | – | MIT | – | Apache-2.0 | MIT |
36 skills (instruction sets) by ai4protein that an AI reads and follows · not an MCP server
Pick the skill to install
Install only the ones you use. Open a skill to see how to install it
This set has 36 skills, too many to install at once: every installed skill’s description loads into every chat, so the AI picks the wrong one more easily. Pick only the ones you use from the list below
- alphafold_database
AlphaFold DB structures and confidence analytics via VenusFactory tools. Use when the user needs predicted structures by UniProt ID, pLDDT/PAE analysis, or PDB/mmCIF download. Do NOT use for experimental PDB (rcsb_database), local ESMFold without UniProt (predict_structure_esmfold / protein_structu…
- arxiv
arXiv preprint server — keyword search the official API and download papers as PDF / HTML / source tarball. Use whenever the user mentions an arXiv ID (e.g. 2106.04559) or wants preprints on a topic in CS / physics / math / quantitative biology. For biomedical preprints specifically, prefer biorxiv…
- biopython
Biopython guidance for sequence I/O, alignments, Bio.PDB, and Entrez parsing. Use for custom bioinformatics code via session Python tools. Prefer VenusFactory download tools for NCBI/UniProt/AlphaFold bulk fetches so large payloads stay on disk. Do NOT use as a substitute for protein_sequence_simil…
- biorxiv
Use for biology and medicine preprints: search bioRxiv or medRxiv and retrieve DOI metadata and version history. Prefer pubmed for peer-reviewed biomedical literature.
- brenda_database
BRENDA enzyme kinetics via VenusFactory download tools (SOAP). Use for Km/kcat, reactions, organism comparison, environmental optima by EC number. Do NOT use for pathway maps alone (kegg_database) or protein sequence fetch (uniprot_database). Requires BRENDA_EMAIL and BRENDA_PASSWORD.
- chembl_database
ChEMBL bioactive molecules and drugs via VenusFactory download tools. Use for molecule/drug by ID, similarity/substructure by SMILES, SAR starting points. Do NOT use for openFDA regulatory data (fda) or RDKit-only local chemistry (rdkit).
- clustalo_msa
Multiple sequence alignment of proteins via EBI Clustal Omega web service. Use when you have ≥2 protein sequences in a FASTA file (≤4000 sequences, ≤4 MB) and need an alignment to assess conservation, residue importance, or domain structure. Do NOT use for: single sequences, homology search (use pr…
- fda
Query openFDA via VenusFactory for drugs, devices, adverse events, recalls, and regulatory submissions (510k, PMA). Use when the user needs FDA pharmacovigilance, labeling, NDC/UNII, or openFDA analytics. Do NOT use for ChEMBL bioactivity (chembl_database) or general biomedical literature (pubmed).
- foldseek_structural_similarity
FoldSeek structural similarity search against PDB with optional protected-region masking. Use when the user has a PDB and wants fold-level homologs, structural neighbors, or to protect an active site while searching. Do NOT use for sequence BLAST/MMseqs2 (protein_sequence_similarity_search) or MSA…
- hpa_expression_context
Human Protein Atlas expression and localization via VenusFactory download tools. Use when the user needs tissue expression, subcellular location, single-cell type, blood expression, or protein summary by gene symbol for therapeutic/target context. Do NOT use for mouse/non-human expression atlases o…
- interpro_domain_annotation
InterPro domain/family annotation via VenusFactory download tools. Use when the user needs domain boundaries, family membership, or UniProt→InterPro annotations for engineering target selection. Do NOT use for pathway enrichment (string_database / kegg_database) or kinetic parameters (brenda_databa…
- kegg_database
KEGG REST access via VenusFactory download tools (academic use). Use for pathway/gene/compound lookups, ID conversion, and DDI. Do NOT use for PPI networks (string_database) or enzyme kinetics (brenda_database). Non-academic use of KEGG requires a commercial license.
- matplotlib
Matplotlib OO/pyplot guidance for custom plots via session Python tools. Use for fine-grained control. Prefer nature_figure for manuscript figures and seaborn for quick statistical EDA.
- nature_figure
Submission-grade Nature/high-impact journal figure workflow for Python or R. Use whenever the user asks to create, revise, audit, or polish manuscript figures, multi-panel scientific plots, figures4papers-style matplotlib plots, or journal-ready SVG/PDF/TIFF outputs, especially for Nature-family or…
- nature_polishing
Use when editing or translating existing scientific prose into Nature-style English, or fixing manuscript LaTeX layout. Prefer nature_writing for drafting new sections from results; not for literature searches.
- nature_writing
Use when planning or drafting scientific manuscript sections from supplied claims, results and figures. Prefer nature_polishing for editing finished prose; not for generating unsupported results.
- ncbi_clinvar
Use for querying NCBI ClinVar variant classifications, review status and supporting evidence by gene or condition. Prefer ncbi_gene for gene annotations; not for diagnosing a patient.
- ncbi_gene
Use for NCBI Gene lookup by symbol or ID, RefSeq accessions, gene annotations and batch queries. Prefer ncbi_clinvar for variant clinical significance and uniprot for protein annotations.
- ncbi_sequence
NCBI E-utilities for biological sequences — fetch protein/nucleotide FASTA by accession, run BLAST, translate CDS to protein, search NCBI Protein by gene+organism. Use when the user provides an NCBI accession (NP_, XP_, NM_, NR_, etc.), asks for a sequence by gene name + species, or needs to transl…
- openalex
Use for scholarly works, authors, institutions and citation networks in OpenAlex. Prefer pubmed for biomedical abstracts and biorxiv for preprint version histories.
- protein_engineering_hypothesis
Evidence-bounded hypothesis and experiment planning for protein engineering. Use when the user asks what to mutate next, how to prioritize variants, how to falsify a mechanism, or how to design a directed-evolution round. Do NOT invent wet-lab results; chain VenusFactory tools for computational evi…
- protein_property_prediction
Physicochemical properties, surface/SS features, and finetuned protein/residue function prediction. Use when the user asks for solubility, optimal temperature, activity/binding/conserved sites, RSA/SASA/secondary structure, or property tables from FASTA/PDB. Do NOT use for zero-shot mutation rankin…
- protein_sequence_similarity_search
Find homologous protein sequences from a query sequence using MMseqs2 (fast, ColabFold web API) or BLAST (comprehensive, EBI). Use when the user provides a protein sequence or FASTA file and wants homologs, function inference by sequence similarity, or input for an MSA. Do NOT use for structural si…
- protein_structure_pipeline
Protein structure obtain → confidence → visualize pipeline. Use when the user needs a 3D structure from sequence or UniProt ID, AlphaFold/ESMFold retrieval, pLDDT/PAE analysis, or structure rendering. Do NOT use for mutation ranking (zero_shot_mutation_workflow), FoldSeek search (foldseek_structura…
- proteinmpnn_design_workflow
ProteinMPNN inverse folding: design or score sequences on a fixed backbone. Use when the user wants sequence design from PDB, interface/binder design, homomer symmetry, or fixed catalytic residues. Do NOT use for zero-shot mutation ranking on a wild-type sequence (zero_shot_mutation_workflow) or de…
- pubmed
Use for biomedical literature searches, PMID citation resolution and structured abstracts in PubMed. Prefer biorxiv for unpublished preprints and openalex for citation networks.
- pymol
Headless PyMOL rendering of protein structures (PNG + PSE session) and structural superposition with RMSD. Use to produce static publication-style images, color a structure by pLDDT/B-factor/chain/secondary-structure, or compare two structures by cealign. Do NOT use for interactive 3D exploration (…
- rcsb_database
RCSB Protein Data Bank (PDB) — experimentally determined 3D biomolecular structures. Search by full-text/sequence/structure/attribute, fetch entry metadata, download coordinate files (PDB/mmCIF). Use when the user provides a PDB ID, asks for structures of a protein, wants to find similar structures…
- rdkit
RDKit cheminformatics via VenusFactory bioinfo scripts and session Python tools. Use for SMILES/SDF, descriptors, fingerprints, substructure filters, similarity. Do NOT use for ChEMBL bioactivity download (chembl_database) or openFDA (fda). No dedicated rdkit_* LangChain tool — run scripts or impor…
- seaborn
Seaborn statistical plots for exploratory analysis via session Python tools. Use for quick relational/distribution/categorical charts. Do NOT use for submission-grade Nature figures (nature_figure) or low-level artists control (matplotlib).
- string_database
STRING PPI networks and enrichment via VenusFactory download tools. Use for interaction networks, partners, GO/KEGG enrichment, homology across 5000+ species. Do NOT use for sequence homology search (protein_sequence_similarity_search) or KEGG pathway entries alone (kegg_database).
- structure_file_prep
Structure/sequence file preparation with VenusFactory file tools. Use for FASTA parsing, PDB chain extraction, PDB↔mmCIF conversion (MAXIT), apo checks, batch PDB→FASTA, and UniProt ID from RCSB metadata. Do NOT use for structure prediction (protein_structure_pipeline) or homology search.
- uniprot_database
UniProt — protein sequence, function, taxonomy, cross-references. Search proteins by query, retrieve a UniProt entry, map IDs between databases (PDB↔UniProt etc.), pull FASTA sequence, fetch metadata, run SPARQL against sparql.uniprot.org. Use whenever the user mentions a UniProt accession (e.g. P0…
- venus_finetune_workflow
Fine-tune and run custom protein models on VenusFactory (CSV/HF → config → train → predict). Use when the user brings labeled sequences, wants adapter training (ProtT5/ESM2/Ankh/QLoRA notes), or batch inference with a trained config. Do NOT use for zero-shot mutation without labels (zero_shot_mutat…
- workflow_skill_creator
Distills a completed user workflow or interaction into a reusable VenusFactory agent skill. Use when the user says "make this a skill", "create a skill from what we just did", "package this workflow" or similar. Adapts the workflow into the VenusFactory tools wiring + SKILL.md pattern. Do not use f…
- zero_shot_mutation_workflow
Zero-shot mutation engineering with VenusFactory PLMs. Use when the user wants beneficial mutations, directed evolution candidates, or stability/fitness ranking from a FASTA sequence or PDB structure. Do NOT use for ProteinMPNN inverse folding (proteinmpnn_design_workflow), sequence homology search…
How a Skill differs from MCP
MCP = the hands
Connects the AI to an outside service or data, such as reading Notion or creating an invoice. A program or server has to be running.
Skill = the manual
A text file that teaches the AI how to do a task. It connects to nothing.
Some overlap: a few Skills run scripts that call tools, and a few MCP servers only send knowledge. Start from the task you want done.
Works with
Not stated by the maker · SKILL.md skills work in apps that support Agent Skills, such as Claude Code, Codex and Cursor
Install
Pick a skill first
This repo has 36 skills. Pick one from the list above; its page shows every way to install it, including letting the AI install it and a ZIP
Discussion
Ask, share a tip or report that it does not work. Sorted by what people marked helpful.
No posts yet. Be the first to ask or share.
Posts here are written by users, not the PopNx team.
From GitHub
- GitHub stars
- 253
Stars are for the whole repo, not each skill - Maker
- ai4protein
- License
- No license stated This repo states no license. By default the author keeps all rights: ask before you share it or use it commercially.
- Last code change
- Nov 18, 2026
- SKILL.md files in the repo (including translated copies and templates)
- 36
- Topics
- agent, database, fine-tuning, language-model, life-science, noob-friendly, protein, protein-science, skill, tutorial