agent-skills

Iterate on an idea with the user, challenging their viewpoint in constructive ways until both parties are content with the outcome. Use when the user wants to develop an idea, refine a half-formed concept, pressure-test their thinking, or mentions "brainstorm".

Computes the Levi-Civita connection, Christoffel symbols, Riemann curvature tensor, Ricci tensor, and scalar curvature from a Riemannian metric using the local_coordinates JAX library. Use when the user works with curvature quantities, parallel transport, Koszul formula, Bianchi identities, or needs to verify that a metric is flat or that curvature tensor symmetries hold.

Solves the geodesic equation and computes exponential and logarithmic maps on Riemannian manifolds using the local_coordinates JAX library. Use when the user works with geodesics, parallel transport along a curve, Taylor expansions of the exponential map in Riemann normal coordinates, injectivity radius, or ODE integration of the geodesic system.

Builds RiemannianMetric objects in the local_coordinates JAX library from metric component functions, including raising and lowering indices, changing basis, and constructing the inverse metric. Use when the user defines a metric tensor, needs to convert between coordinate and orthonormal bases, or wants to verify symmetry and positive definiteness.

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me".

Use this skill when constructing, training, evaluating, saving, loading, or documenting the JAX plus Equinox conformal coordinate frame field for disentanglement, a noise-conditional model that learns a local conformal frame J(x) = lambda(x) U(x) from data through score matching, integrability, and independence losses.

Use this skill when constructing, applying, training, saving, loading, or documenting a standalone learnable orthogonal matrix in JAX plus Equinox, parameterized by the matrix exponential of a skew-symmetric matrix, the Cayley transform, or a QR factorization.

Use this skill when constructing, applying, training, saving, loading, or documenting a network-predicted orthogonal transformation on vectors in JAX plus Equinox, where a small residual MLP predicts the parameters of an orthogonal matrix from an input, using the matrix exponential, the Cayley transform, or a QR factorization.

Use this skill when constructing, sampling, training, saving, loading, or documenting the JAX plus Equinox StyleGAN3 implementation from the stylegan3_jax repository.

Use this skill when constructing, sampling, training, saving, loading, or documenting the JAX plus Equinox port of TarFlow, the autoregressive transformer normalizing flow from Apple's ml-tarflow paper.

Uses second-order Jets (value, gradient, Hessian triples) for automatic differentiation in the local_coordinates JAX library, including pushforward through smooth maps, change of coordinates, and the @jet_decorator pattern. Use when the user works with Jets, second-order Taylor expansions, JVPs and HVPs, or needs to propagate derivatives through a composition of maps.

Emits a LeJEPA encoder (backbone, 3-layer MLP projector with BatchNorm, and SIGReg loss as separate objects) in PyTorch or equinox+JAX, mirroring the MINIMAL.md assembly. Bring your own backbone is the primary API; a default backbone is included (timm ViT-S/8 for PyTorch, a minimal self-contained equinox ViT for JAX). Use when the user wants the three pieces as separate objects and intends to write their own training loop. For a single bundled object, use `lejepa-model` instead.

Emits the SIGReg (Sketched Isotropic Gaussian Regularization) loss module from LeJEPA in PyTorch or equinox+JAX. SIGReg is a sliced characteristic- function test for standard Gaussianity. Use when the user wants only the SIGReg loss as a drop-in module, references LeJEPA's regularizer, or needs to add SIGReg to an existing pretraining pipeline. For the full LeJEPA recipe (backbone + projector + invariance), use `lejepa-model` or `lejepa-encoder` instead.

Emits a bundled `LeJEPAModel` class in PyTorch or equinox+JAX with `embed(x)` for downstream evaluation and `loss(views)` for training. The class owns backbone, projector, and SIGReg internally, and exposes lambda as a constructor argument. The online linear probe lives outside the model. Use when the user wants a single-object LeJEPA API. For separately wired pieces (backbone, projector, SIGReg as distinct objects), use `lejepa-encoder` instead.

Emits a single-GPU LeJEPA pretraining script in PyTorch or equinox+JAX, matching MINIMAL.md. Includes AdamW with separate weight decay groups, linear warmup plus cosine annealing schedule, bf16 mixed precision, the lightweight V-augmented-views augmentation pipeline, and an optional online linear probe controlled by a flag. Use when the user wants a runnable pretraining script with paper-default hyperparameters. Does not ship DDP, checkpointing, or full DINO multi-crop (those are documented as upgrade paths in the reference).

Entry point for LeJEPA (Lean Joint-Embedding Predictive Architecture) and SIGReg (Sketched Isotropic Gaussian Regularization). Explains the paper's recipe (backbone, projector, SIGReg loss, invariance loss with a single trade-off lambda) and the four child skills that instantiate it at different scopes in either PyTorch or equinox+JAX. Use when the user mentions LeJEPA broadly, asks about the paper, is unsure which child skill applies, or wants background before picking a scope.

Entry point for the local_coordinates JAX library for differential geometry computations, including Riemannian metrics, connections, curvature, geodesics, pullback metrics, Riemann normal coordinates, and second-order Jet differentiation. The library is pip-installable from https://github.com/EddieCunningham/local_coordinates. Use when the user asks about the library as a whole, needs an orientation, or is unsure which of the related skills applies to their task.

Computes pullback metrics under smooth maps and coordinate changes using the local_coordinates JAX library, including Jacobian-based component formulas and general (k,l) tensor transformation laws. Use when the user applies a coordinate change to a metric, pulls a metric back through an immersion or a learned map, or works with induced metrics on submanifolds.

Constructs Riemann normal coordinates (RNC) at a point on a Riemannian manifold and transforms geometric objects into them using the local_coordinates JAX library. Use when the user centers a computation at a point, exploits the fact that the metric is the identity with vanishing first derivatives at the origin, or needs Taylor coefficients of the exp/log map in terms of the Riemann tensor.

Writes and reviews Slurm sbatch scripts for GPU clusters including gres GPU requests, partitions, memory and CPU sizing, time limits, job arrays, logging, multi-GPU srun patterns, the discipline of only requesting a GPU when the workload genuinely uses one, and the preflight/sampling/audit pattern that verifies a GPU job actually used the GPU. Use when the user mentions sbatch, Slurm, HPC, GPU jobs, Unity or other campus clusters, preempt GPU partitions, hyperparameter sweeps, or batch submission of ML experiments.