One Forge, All Skills: A curated skill collection for academic writing and research. 点开即用,按需配置的一站式学术研究skills平台。
-
Updated
Aug 30, 2026 - Python
One Forge, All Skills: A curated skill collection for academic writing and research. 点开即用,按需配置的一站式学术研究skills平台。
A framework for discovering, compiling, and validating reusable skills for scientific agents.
Learning to Evaluate Before Improving: Automatic Rubric Induction for Automatic Research Agents
Scientific Agent for evidence-grounded evaluation of cell-therapy products.
Agent-native, local-first plugin for auditable single-cell annotation with governed evidence, contrastive review, conservative abstention, and human-reviewed draft outputs.
Evidence-governed autonomous experimentation and falsification for computational science
An auditable human–AI workflow from NNDC/ENSDF data to gamma-ray GCD lifetime inference.
A falsification-first protocol for scientific AI agents: claim contracts, hidden validation, decoys, calibrated scoring, and independent evidence.
Trustworthy biomedical tool-using agent with task-conditioned planning, execution-grounded verification, and traceable repair.
A verifier-gated research platform for skill transfer in scientific agents for systems biology.
Auditable chemistry skills for scientific agents with deterministic computation, provenance, and explicit scientific boundaries.
Open Earth remote-sensing MCP server that lets AI agents discover geospatial data and run tools for STAC, Sentinel-2, nightlights, water mapping, spectral indices, and SAR metadata.
Auditable SciForge × BioGym de novo protein-design run: RFdiffusion → ProteinMPNN → Boltz-2
Evidence-grounded scientific agents with replayable evaluation and human review — the public integration + evaluation layer for AI-for-science workflows.
Open agent runtime, Skills, evidence standards, and instrument plugins for experimental science.
Causal evidence-uptake stress tests and an Epistemic Brake for scientific agents.
Organization profile and community standards for the SciPhys open agent stack.
Process traces for evaluating AI scientist workflows | ICML 2026 AI4Science Dataset Competition | 432 trajectories from GPT-5.4, Claude Opus 4.6, Gemini 3.1 Pro
Auditable MCP tools for PyLabRobot: run and sample context, typed provenance, simulation by default, and explicit hardware permission.
验证边界 · Verification Frontier — 安全扩展科学 Agent 的验证边界 / Safely Engineering the Scientific Agents. GOAI 2026 AI for Research · Open Exploration. Bilingual 中文+EN, no-login demo, deterministic reproducible package.
To associate your repository with the scientific-agents topic, visit your repo's landing page and select "manage topics."