AI Agent Workflow Collection

Use the InfiniSynapse CLI (agentinfini) for multi-turn AI data-analysis tasks, database/RAG context, and task workspace files. Use when the user mentions InfiniSynapse, agentinfini, database or RAG analysis, or asks to delegate analysis through InfiniSynapse.

Predict protein structure for monomers and multimers with AlphaFold2 via the ColabFold runner (Mirdita et al. 2022, github.com/sokrypton/ColabFold; AlphaFold2 Jumper et al. 2021). Reach for this skill to fold a sequence or complex with the AF2/AF2-Multimer evoformer, to validate designed sequences by self-consistency pLDDT, ipTM, and RMSD, or to run a quick MSA-backed prediction using the public MMseqs2 server.

Organize multi-step scientific analyses into reproducible, self-contained modules. Use for workflows such as QC→PCA→DEG→GSEA that produce scripts, inputs, figures, tables, and methods. Creates a stable module layout, records exact inputs/parameters/package and database versions in each module README, keeps large data as references instead of copies, and verifies outputs before completion.

Predict genome-wide functional tracks (RNA-seq, CAGE, DNase, ChIP) from DNA sequence with Borzoi. Use this skill when: (1) Scoring the regulatory effect of a variant on expression/accessibility, (2) Generating predicted coverage tracks for a locus, (3) Prioritising non-coding variants by predicted track delta.

Structure prediction for protein, nucleic-acid, and small-molecule complexes with the Chai-1 foundation model (Chai Discovery 2024, github.com/chaidiscovery/chai-lab). Reach for this skill to predict an antibody-antigen or protein-ligand complex from a single FASTA, to re-fold designed binders as an AlphaFold-multimer alternative, or to drive co-folding from Python for batched campaigns on a GPU.

Set up a compute environment on a remote provider so wisp-science jobs can run there. Covers direct SSH/conda hosts, Slurm clusters, container-via-bridge runners, and managed-API providers (Modal, GCP, RunPod). Use when standing up a new provider, porting an env to a different backend, adding a tool that needs its own software stack, or wiring weight caches. Triggers on "new compute provider", "set up env on", "port env to", "build GPU image", "weight cache", "computedetails", "conda env on the box", "apptainer on slurm".

Biohub ESMFold2 / ESMFold2-Fast all-atom co-folding (Candido et al. 2026, github.com/Biohub/esm). Single-sequence and MSA modes; protein, DNA, RNA, ligand (CCD/SMILES), modified residues. FoldBench Ab-Ag 50-55%, PPI 70-77% DockQ-pass. Also covers the ESMC-{300M,600M,6B} protein language models from the same release: masked-LM logits, hidden states, mutation scoring, contact prediction, and the SAE interpretability head. MIT-licensed weights on HuggingFace org biohub. Use this skill when: (1) Predicting complex structures with single-sequence input, (2) Validating designed binders with ESMFold2-Fast, (3) Running ESMFold2 with MSA input, (4) Getting ESMC embeddings or per-residue mutation scores, (5) Choosing kernel backend and sampling-step settings for paper-faithful throughput.

Embed proteins with Meta AI's ESM-2 (fair-esm package). Use this skill when: (1) Extracting per-residue or per-sequence embeddings for downstream ML, (2) Masked-LM likelihood / mutation effect scoring, (3) Contact prediction from a sequence.

Compose one publication-grade multi-panel figure. Entry from a one-line claim + data refs, OR from an existing figure via deriveoutline(png). Runs a per-figure loop: outline (12-col grid, per-panel ask + labelbudget) → fan-out one sub-agent per panel (each loads figure-style) → tile + stamp letters → adversarial composite review with two-tier feedback (Tier-1 outlinerevisions / Tier-2 per-panel violations) → regen affected panels, ≤3 rounds. Loads paneltask / composefigure / composecrops / compositereviewtask / deriveoutline into the kernel. For one standalone plot use figure-style; for whole-paper figure ordering use paper-narrative.

Publication-grade figure correctness and legibility rules. Load before drawing any plot and call applyfigurestyle() — sets a role-mapped font-size ladder, outward ticks, frameless legends, and 300-dpi output. The skill is a checklist, not a house look: data fidelity (claim-titles tested against every row, excluded data never enters summaries), label economy (floor and ceiling), colour threading, chart-choice-by-data-shape, layout, and a render-then-verify QA loop (bbox collision + per-panel perceptual check). Ships helpers: focalpalette, barwithpoints, stripwithmedian, endoflinelabels, panelletter, setframe, panelcrops. For multi-panel figures load figure-composer; for whole-paper figure arc load paper-narrative.

Use when the user provides a scientific paper PDF and asks for a group-meeting literature report, journal-club slides, 文献汇报PPT, 组会PPT, paper presentation, article walkthrough, or to explain a paper with PowerPoint. The skill first reconstructs the paper's scientific logic, then builds author/background sections, chooses an evidence-driven slide outline, crops only main-text figure panels from the PDF, and creates an academic PPT with 10–30 slides including title and conclusion/discussion. Always use this skill for '把这篇论文做成PPT', '文献汇报', 'journal club', or '组会汇报', even if the user only says they need slides.

Configure the local wisp-science runtime — uv/Python bootstrap, Node+scimaster-cli for bear- literature skills, pixi for bioinformatics multi-env analysis. Detect mainland-China network and apply mirrors. Use when Capabilities shows missing Python/uv/Node/sci/pixi, bootstrap errors, or the user asks to 配置环境 / install Python / uv / Node / pixi / set up the local environment. Not for remote GPU/SSH compute (use compute-env-setup).

Register a model service in the managed family — a local model server container the daemon starts/stops on demand, or a remote upstream model API (https). Read the runbook, allocate a port (local only), compose idempotent start/stop scripts (local only), register once. Load when the user wants a model service available for inference, or when listcompute shows managed endpoints.

Structure prediction using OpenFold3, an open-weights PyTorch reproduction of AlphaFold3 from the AlQuraishi Lab. Use this skill when predicting protein/nucleic-acid/ligand complex structures with an Apache-2.0-licensed AF3 reimplementation.

Judge and reshape the STORY a paper's figures tell. Input is the work itself — manuscript (or abstract) + figure deck — no hand-written brief. derivepaperbrief(abstract, captions) extracts pitch/vision/per-figure-claims; a handling-editor reviewer on the full deck returns hookverdict (would Fig 1 make me send this for review?), arc (hook→mechanism→evidence→application), figuremoves (panels in the wrong figure), missingpanels (concrete analyses to RUN), killlist, and boldestdefensiblefig1. Hands per-figure claims to figure-composer. Load when writing or revising a paper.

Use when the user has attached a PDF, paper, report, or other document and the answer needs content from more than one place in it: summarize the methods or any other section, compare sections, find where a topic is discussed, read a value or label off a figure or chart, or find/list/extract every instance of something across the whole document (datasets, benchmarks, citations, figures, table rows, accession numbers — including appendices). Skip it only for a single lookup of 1–4 pages quoted in your very next response — readfile(pages=[...]) attaches pages as images that are dropped from context after one turn, so multi-section answers end up re-reading the same ranges repeatedly. Parses the PDF once in the Python kernel: pdfpages (pages as persistent text), pdfoutline (TOC), pdfscan (rank pages by relevance), pdfmap/pdfextract (per-page summary / structured fields via parallel haiku calls). For PDF creation/manipulation, use reportlab/pypdf directly.

Inspect a registered execution server before compute planning and interpret its persisted capability profile. Use when a server is added, when the user clicks Probe, before enabling an unfamiliar SSH/WSL resource, or when deciding whether GPU, sudo/root, a scheduler, Python, R, conda, mamba, or environment modules are available.

Run GPU jobs on the user's own Modal account via host.compute.create('byoc:modal', ...). Covers the create→submit→waitfornotification flow, the computeprovider kernel for env setup, image/volume resolution, and the two approval cards. Load once you've decided to dispatch to Modal.

Submit recoverable SSH-direct research Runs through Wisp's local control plane without blocking the conversation.

Embed and annotate single-cell expression data with scGPT, a foundation model for single-cell biology. Use this skill when: (1) Producing cell embeddings from an AnnData for clustering/integration, (2) Zero-shot or fine-tuned cell-type annotation, (3) Gene-level representation for perturbation/GRN tasks.

Call a registered model endpoint over its native HTTP API from the endpoint's scoped inference kernel (BASEURL preloaded). Load once a task needs predictions from a registered model endpoint.

给一个概念,从真实检索到的论文摘要里挖出它的邻近概念,画出以该概念为中心的知识地图——每个节点都锚定真实文献,不从记忆里补节点。同时输出 Mermaid 概念图(可在 Claude Code / GitHub / Obsidian 渲染)和可截图的独立 HTML 地图,再给 3–6…