Draft, revise, and audit scientific manuscripts with enforced evidence provenance, reporting-guideline coverage, and authorship/disclosure checks.
A rigorous skill for validating, converting, and symmetry-analyzing crystal structures with current pymatgen APIs, plus offline phase diagrams and bounded Materials Project queries.
An end-to-end PyMC workflow skill covering hierarchical models, NUTS/variational inference, convergence diagnostics, and LOO/WAIC model comparison.
Run end-to-end proteomics and metabolomics LC-MS/MS analysis with ready-made pyOpenMS CLI scripts.
A skill that guides Claude through choosing SHAP explainers and maskers, computing, validating, and visualizing feature attributions on SHAP 0.52.
A fail-closed EDA skill that profiles authorized local scientific files, audits missingness, leakage and outliers, and drafts a rigorous report without exposing raw values.
Design, review, and migrate MATLAB or Octave code with strict license gating, static-only safety checks, and reproducibility reporting.
Build, test, and analyze bounded process-based discrete-event simulations in SimPy with correct event semantics and replication-based output analysis.
A hands-on PyG skill covering node/link/graph tasks, custom message passing, heterogeneous graphs, and neighbor sampling for large graphs.
Builds macro-free, editable PowerPoint conference posters from an author-approved local manifest, with physical-size, printer, accessibility, and package-security audits.
A skill for designing, simulating, and running quantum circuits with Google's Cirq framework.
Correct, idiomatic pysam workflows for reading, querying, filtering, and writing SAM/BAM/CRAM, VCF/BCF, FASTA/FASTQ, and tabix data.
Turns observations into evidence-bounded candidate hypotheses, rival explanations, discriminating predictions, and preregistration-ready analysis plans.
A Remotion-based pipeline that turns a frontend project or webpage into a cinematic promo video using real page screenshots, 2.5D camera moves, beat-synced cuts, and sound design.
A P0–P7 research pipeline with parallel subagents, a locked citation registry, source-type governance and mandatory counter-review for traceable reports.
Picks a multi-agent orchestration pattern from requirements, generates and validates provider-ready tool schemas, and evaluates run logs for cost, latency, and bottlenecks.
Compiles a goal into a verifiable task plan and drives a bounded agent loop with machine-run checks, retry caps, and human escalation.
Spawns N parallel subagents on the same task in isolated git worktrees, ranks their results, and merges the winning branch.
Installs and configures the GitHub channel adapter so a NanoClaw agent can converse inside PR and issue comment threads.
Spins up five mutually hostile expert agents, runs a three-round cross-critique debate, and hands only the surviving insights to a dedicated plan agent for an executable work plan.