InternAgent
InternAgent-1.5: A Unified Agentic Framework for Long-Horizon Autonomous Scientific Discovery
- 1.5k
- Python
- 4 mo ago

The maintainer has not written a description at the source.
Repo InternScience/scp, read from GitHub 3 d ago. InternScience has not joined Take My Repo.
InternScience/scp is a public repo on GitHub by InternScience. It declares the MIT license on GitHub. It is written mainly in Jupyter Notebook, has 173 stars and last changed on 2026-06-03. Nobody has reviewed it on Take My Repo yet, which says nothing about its quality. Read from GitHub on 2026-11-18.
In the maintainer's words. Shown from GitHub; Take My Repo does not keep a copy.
Science Context Protocol โ Building a Global Collaboration Network for Autonomous Scientific Agents
SCP is dedicated to breaking down tool barriers in scientific research, enabling researchers to focus on the science itself, and empowering AI agents to become true "experiment designers."
SCP (Science Context Protocol) is an open-source standard protocol designed to accelerate scientific discovery by building a global collaboration network for autonomous scientific agents, connecting heterogeneous scientific resources (software tools, AI models, datasets, workflow engines, lab instruments, etc.).
| Capability | Description |
|---|---|
| ๐ Protocol-Level Connectivity | Unified description and invocation of 2,200+ scientific resources (tools, models, instruments, etc.) |
| ๐ง Intelligent Orchestration | SCP Hub supports automated planning, execution, and monitoring of multi-step workflows |
| ๐งช Dry-Wet Lab Integration | Seamless integration of computational tools and lab equipment for closed-loop validation |
| ๐ค Multi-Agent Collaboration | Supports multiple AI agents working collaboratively within a unified context |
| ๐ Full Lifecycle Management | End-to-end traceability from registration, planning, and execution to archiving |
| ๐ Security & Access Control | Fine-grained authentication and authorization mechanisms based on experiments |
SCP adopts a Hub-Spoke Architecture:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ SCP Client โ
โ (Researchers / AI Scientists) โ
โโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ
โผ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ SCP Hub (Central Orchestrator) โ
โ Intent Parsing ยท Workflow Generation ยท Task Scheduling โ
โ ยท Permission Management โ
โโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ
โโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโ
โผ โผ โผ
โโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโ
โ SCP Server โ โ SCP Server โ โ SCP Server โ
โ (Edge Node) โ โ (Edge Node) โ โ (Edge Node) โ
โโโโโโโโฌโโโโโโโโ โโโโโโโโฌโโโโโโโโ โโโโโโโโฌโโโโโโโโ
โ โ โ
Tools/Models Databases/ Compute
Instruments Resources
# Clone the repository
git clone https://github.com/InternScience/scp.git
cd scp
# Install dependencies
pip install -e .
Set up your own SCP Server and Hub for full control over deployment and management.
Visit SCP Square โ we provide a managed SCP Hub where you can submit your own SCP Server to the square for cross-platform discovery and access.
SCP has integrated 2,200+ scientific tools spanning multiple disciplines:
| Domain | Share | Representative Tool Types |
|---|---|---|
| ๐งฌ Biology & Related Technologies | 45.9% | Genome analysis, protein structure prediction |
| โ๏ธ Physics | 21.1% | Quantum computing, materials simulation |
| ๐งช Chemistry | 11.6% | Molecular docking, reaction prediction |
| ๐ง Mechanics & Materials Science | 8.7% | Finite element analysis, molecular dynamics |
| ๐ Mathematics | 8.0% | Symbolic computation, optimization algorithms |
| ๐ป Information Science & Computing | 4.6% | Machine learning, data mining |
Building upon tool standardization, SCP further introduces the concept of "Scientific Skills (SCP Skills)" โ the first batch of 206 scientific skills is now available in the
skills/directory.
Each Skill is an independently callable, freely composable "research instruction" that orchestrates multiple databases, computational models, and even device drivers behind the scenes, forming a reusable, standardized experiment template.
For Researchers: No need to deal with complex tool manuals and code debugging โ focus your energy on the scientific questions themselves For AI Agents: Skills become freely stackable "LEGO bricks," enabling agents to evolve from "tool users" to "experiment designers"
All 206 skills are organized into 8 major scientific domains:
| Domain | Skills | Representative Examples |
|---|---|---|
| ๐ Drug Discovery & Pharmacology | 71 | Target identification, ADMET prediction, virtual screening, molecular docking, drug safety & repurposing, clinical pharmacology |
| ๐งฌ Genomics & Genetic Analysis | 41 | Variant pathogenicity, cancer genomics, population genetics, rare disease, virus genomics, epigenomics |
| ๐งฌ Protein Science & Engineering | 38 | Structure prediction (ESMFold/AlphaFold), binding sites, mutation analysis, antibody & peptide design, enzyme engineering |
| ๐งช Chemistry & Molecular Science | 24 | Structure analysis, molecular fingerprints, SAR, material composition, natural products, metabolomics |
| โ๏ธ Physics & Engineering Computing | 18 | Circuit analysis, thermodynamics, optics, electromagnetics, crystallography, unit conversion |
| ๐ฌ Lab Automation & Literature Mining | 7 | Protocol generation, PDF extraction, PubMed search, scientific literature mining, meta-analysis |
| ๐ Earth & Environmental Science | 5 | Atmospheric science, wind energy assessment, seawater properties, oceanographic calculations |
| ๐ Biomedical Database Integration | 2 | OpenTargets disease-target queries, experimental data processing |
admet_druglikeness_report ยท affinity_maturation ยท alanine_scanning_pipeline ยท antibody_drug_development ยท atc_drug_classification ยท boltz2-binding-affinity ยท cancer_therapy_design ยท chemical_safety_assessment ยท clinical_pharmacology_report ยท clinical_trial_drug_profile ยท combinatorial_chemistry ยท comparative_drug_analysis ยท compound_to_drug_pipeline ยท disease-reversal-prediction ยท disease_compound_pipeline ยท disease_drug_landscape ยท disease_knowledge_graph ยท drug-screening-docking ยท drug_indication_mapping ยท drug_interaction_checker ยท drug_metabolism_study ยท drug_repurposing_screen ยท drug_safety_profile ยท drug_target_identification ยท drug_target_structure ยท drug_warning_report ยท drugsda-admet ยท drugsda-compound-retrieve ยท drugsda-data-valid ยท drugsda-denovo-sampling ยท drugsda-dleps ยท drugsda-drug-likeness ยท drugsda-esmfold ยท drugsda-file-transfer ยท drugsda-linker-sampling ยท drugsda-mol-properties ยท drugsda-mol-similarity ยท drugsda-mol2mol-sampling ยท drugsda-p2rank ยท drugsda-peptide-sampling ยท drugsda-prosst ยท drugsda-rgroup-sampling ยท drugsda-target-retrieve ยท enzyme_inhibitor_design ยท epigenetics_drug ยท fda-drug-risk-assessment ยท gene_to_drug_pipeline ยท gene_variant_drug_nexus ยท infectious_disease_analysis ยท lead_compound_optimization ยท molecular_docking_pipeline ยท one_health_analysis ยท opentargets-disease-target ยท orphan_drug_analysis ยท pandemic_preparedness ยท pediatric_drug_safety ยท personalized_medicine ยท pharmacogenomics_analysis ยท pharmacokinetics_profile ยท polypharmacology_analysis ยท precision_oncology ยท protein_drug_interaction ยท smiles_comprehensive_analysis ยท structural_pharmacogenomics ยท substance_toxicology ยท systems_pharmacology ยท toxicity_assessment ยท variant-pharmacogenomics ยท virtual_screening ยท gene-knowledge-integration ยท molecular-property-profiling
biomarker_discovery ยท biosample_genomics ยท chromosome_analysis ยท comprehensive-variant-annotation ยท cross_species_genomics ยท epigenomic_landscape ยท gene_disease_association ยท gene_expression_atlas ยท gene_family_evolution ยท genetic_counseling_report ยท genome_annotation ยท go_term_analysis ยท microbiome_genomics ยท mouse_model_analysis ยท multispecies_gene_analysis ยท ncbi-gene-retrieval ยท ncbi_gene_deep_dive ยท organism_classification ยท phenotype-by-hpo-id ยท population_genetics ยท rare_disease_genetics ยท region-gene-elements ยท regulatory_region_analysis ยท snp_functional_analysis ยท tcga-gene-expression ยท tissue_specific_analysis ยท transcriptome_analysis ยท ucsc_genome_exploration ยท variant-clinical-significance ยท variant-cross-database-ids ยท variant-functional-prediction ยท variant-genomic-location ยท variant-gwas-associations ยท variant-population-frequency ยท variant_pathogenicity ยท virus_genomics ยท gene_comprehensive_lookup ยท gene_therapy_target ยท ensembl-sequence-retrieval ยท kegg-gene-search ยท multiomics_integration
alphafold_structure_pipeline ยท antibody_target_analysis ยท binding_site_characterization ยท blast_protein_analysis ยท comprehensive-protein-analysis ยท disease_protein_profiling ยท dna-rna-sequence-analysis ยท enzyme_engineering ยท full_protein_analysis ยท interproscan-domain-analysis ยท interproscan_pipeline ยท metabolomics_pathway ยท molecular_visualization_suite ยท mutation_impact_analysis ยท peptide-properties-calculation ยท protein-blast-search ยท protein-properties-calculation ยท protein_classification_analysis ยท protein_complex_analysis ยท protein_database_crossref ยท protein_engineering ยท protein_function_annotation ยท protein_interaction_network ยท protein_property_comparison ยท protein_quality_assessment ยท protein_similarity_search ยท protein_solubility_optimization ยท protein_structure_analysis ยท proteome_analysis ยท string-ppi-enrichment ยท structural_homology_modeling ยท synthetic_biology_design ยท uniprot-protein-retrieval ยท uniprot_deep_analysis ยท lab_protocol_from_literature ยท code_execution_analysis ยท molecular-descriptors-calculation ยท web_literature_mining
aliphatic_ring_analysis ยท bioassay_analysis ยท cas_compound_lookup ยท cell_line_assay_analysis ยท chembl-molecule-search ยท chemical-mass-percent-calculation ยท chemical-structure-analysis ยท chemical_patent_analysis ยท chemical_property_profiling ยท chemical_structure_comparison ยท compound-name-retrieval ยท compound_database_crossref ยท functional_group_profiling ยท molecular-format-conversion ยท molecular-properties-calculation ยท molecular-similarity-search ยท molecular_fingerprint_analysis ยท natural_product_analysis ยท polymer_property_analysis ยท pubchem-smiles-search ยท pubchem_deep_dive ยท smiles-to-cas-conversion ยท substructure_activity_search ยท material-density-volume-calculation
buoyancy-acceleration-calculation ยท capacitance-calculation ยท electrical_circuit_analysis ยท electromagnetic_analysis ยท energy_conversion ยท experimental_data_processing ยท geometric-volume-calculation ยท geometry_trigonometry ยท length_measurement ยท measurement-error-analysis ยท mobility_analysis ยท nuclear_physics ยท optical-frequency-calculation ยท optics_analysis ยท signal_processing ยท statistical_error_analysis ยท thermal_analysis ยท unit_conversion_suite
biomedical-web-search ยท meta-analysis-execution ยท protocol-extraction-from-pdf ยท protocol-generation ยท protocol-to-executable-json ยท pubmed-article-search ยท scientific-literature-search
atmospheric-science-calculations ยท oceanographic-seawater-properties ยท seawater-freezing-temperature ยท seawater-sound-speed-calculation ยท wind-site-assessment
seismic-waveform-processing ยท unit-conversion-nanoscale
SKILL.md with documentation, test cases, and runnable Python code<YOUR_SCP_HUB_API_KEY> with your own key applied and generated in each SCP server page listed in SCP Square๐ก Want more? If you'd like any useful models or tools to be available as 24/7 accessible APIs, just let us know by submitting an issue, and we'll deploy them for you.
Please refer to the SCP Skills Tutorial.
| Scenario | Description |
|---|---|
| ๐ Automated Experiment Protocol Design | Generate executable experiment protocols from natural language objectives |
| ๐ PDF Protocol Auto-Reproduction | Extract experiment steps from PDFs and execute them automatically |
| ๐ AI-Driven Molecular Screening | Integrated QED scoring, ADMET prediction, and molecular docking |
| ๐งช Dry-Wet Integrated Protein Engineering | Closed-loop workflow from sequence design to experimental validation |
๐ For detailed case studies, please refer to our Technical Report or User Cases
| Feature | MCP | SCP |
|---|---|---|
| Protocol Standardization | General tool invocation | Structured, complete scientific experiment workflows |
| High-Throughput Experiment Support | No built-in experiment management | Supports batch experiments and context management |
| Multi-Agent Collaboration | Point-to-point communication | Centralized orchestration and task distribution |
| Wet-Lab Equipment Integration | Requires custom adapters | Standardized device drivers and interfaces |
| Resource | Link |
|---|---|
| ๐ ๏ธ SCP Tool Square | Explore 2,200+ integrated tools |
| ๐ SCP Skills Tutorial | Docker setup, model switching & skills usage guide |
| ๐ Documentation (Chinese) | Detailed user guide |
| ๐ Technical Report | SCP design and experiment details |
| ๐ฌ Community | Discussions and Q&A |
This project is open-sourced under the Apache License 2.0.
SCP is developed by Shanghai AI Laboratory. We thank the open-source community for their support.
If you use SCP in your research, please cite our technical report:
@article{jiang2025scp,
title={SCP: Accelerating Discovery with a Global Web of Autonomous Scientific Agents},
author={Jiang, Yankai and Lou, Wenjie and Wang, Lilong and Tang, Zhenyu and Feng, Shiyang and Lu, Jiaxuan and Sun, Haoran and Pan, Yaning and Gu, Shuang and Su, Haoyang and others},
journal={arXiv preprint arXiv:2512.24189},
year={2025}
}
Made with โค๏ธ by Shanghai AI Laboratory
InternScience/scp is a public repo on GitHub by InternScience. Its maker has not written a description.
It declares the MIT license on GitHub.
Its maker has not tagged a system, so it is unknown here. Its README says what it needs.
Open "Take it" on its page: Take My Repo writes an install plan for your machine and hands it to your AI agent (Claude Code, Codex or Cursor), which asks you before each step. Take My Repo has not tested it on any machine.
Its last change on GitHub was on 2026-06-03. Take My Repo does not judge from that whether it is maintained.
Nobody has reviewed it on Take My Repo yet, which says nothing about its quality.
InternAgent-1.5: A Unified Agentic Framework for Long-Horizon Autonomous Scientific Discovery
GraphGen: Enhancing Supervised Fine-Tuning for LLMs with Knowledge-Driven Synthetic Data Generation

An open, curated collection of Agent Skills for scientific research โ clone it, use it, extend it!
MLEvolve is an open-source autonomous system for end-to-end machine learning algorithm design and optimization powered by progressive search and experience-driven memory.

A curated collection of papers, datasets, and resources on Scientific Datasets and Large Language Models (LLMs)
Take My Repo has not tested this on your machine. Requirements come from the project's own documentation.