By borisbolliet
classy_szlite assistant: pure-JAX cosmology (CMB Cls, Pk, distances, derived params, halo-model tSZ Cl^yy) backed by v2 CosmoPower emulators. Pipelines, gradient probes, NUTS / SBI / Fisher recipes, cobaya bandpower likelihood scaffolding.
classy_szlite — pure-JAX cosmology code (CMB Cls, matter Pk linear/nonlinear, distances, derived params, halo-model tSZ Cl^yy). Backed by the v2 CosmoPower emulators (same as ACT DR6 extended-cosmology analyses). Differentiable via jax.grad, JIT-friendly, and built around a fast-path `cl_yy_factory` closure for fixed-cosmology MCMC over profile parameters. Use when writing or debugging classy_szlite calculations, scaffolding a tSZ Cl^yy bandpower likelihood, running cobaya MCMCs, or using NUTS / HMC samplers on the GNFW pressure profile.
Compute and plot the tSZ C_ell^yy quickly using classy_szlite (cl_yy_factory). For sanity checks, parameter sweeps, gradient probes, and benchmarks before committing to a long MCMC run.
Scaffold a cobaya likelihood for a tSZ Cl^yy bandpower dataset using classy_szlite. Produces a Python file (Likelihood + Theory + Foreground Theory) and a working cobaya YAML, ready for cobaya-run. The Theory uses classy_szlite.cl_yy_factory so per-step cost is ~5 ms/eval after init.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Specialized Claude Code assistance for classy_szlite — a pure-JAX cosmology code that exposes:
…all backed by the high-accuracy v2 CosmoPower emulators used in
the ACT DR6 extended-cosmology analyses,
and entirely differentiable with jax.grad / jax.jacfwd.
Designed to compose with the cobaya-claude-plugin — enable both to get a full tSZ MCMC workflow.
/class-sz:explain — auto-loads when the conversation is about
classy_szlite, tSZ Cl^yy bandpower fits, GNFW / Arnaud pressure
profiles, the cl_yy_factory fast path, JAX gradient probes, or
cobaya wiring of a classy_szlite-backed theory. Bundles a focused
classy_szlite API + parameter reference.
/class-sz:tszfast [param=value …] — runs a quick Cl^yy
calculation through classy_szlite.cl_yy_factory; supports
parameter sweeps and jax.grad probes. Useful for sanity-checks
before a long MCMC.
/class-sz:build-likelihood <Name> [--data-dir DIR] — scaffolds
a standalone Gaussian likelihood (no SOLikeT dep) + classy_szlite
Theory + foreground Theory + a working cobaya YAML for a tSZ
Cl^yy bandpower dataset (binned bandpowers + N×N covariance — not
a y-map pixel likelihood). Uses cl_yy_factory so the per-step
cost is ~5 ms/eval after init.
class-sz-engineer subagent — specialist for end-to-end
classy_szlite + cobaya work (scaffold, install, run, summarise).
Use for heavy multi-step tasks where you don't want install / chain
output flooding the main thread.
In any Claude Code session, run these three commands:
/plugin marketplace add https://github.com/borisbolliet/class-sz-claude-plugin.git
/plugin install class-sz@class-sz-claude-plugin
/reload-plugins
After /reload-plugins, /class-sz:explain, /class-sz:tszfast,
/class-sz:build-likelihood show in /help and the
class-sz-engineer subagent appears in the Agent picker.
To update later (after I push a new commit):
/plugin uninstall class-sz@class-sz-claude-plugin
/plugin marketplace remove class-sz-claude-plugin
/plugin marketplace add https://github.com/borisbolliet/class-sz-claude-plugin.git
/plugin install class-sz@class-sz-claude-plugin
/reload-plugins
The skills and the class-sz-engineer subagent assume:
~/pyvenvs/py312-class_sz/bin/python (with
classy_szlite, cobaya, numpyro, getdist, jax)~/GitHub/classy_szlite~/class_sz_data/ede/ (or the path in
$CLASSY_SZLITE_DATA_DIR) — see
cosmopower-organization/ede~/Desktop/class-sz-plugin-tests/ (data/, clyy_v2.py, clyy_v2.yaml,
may26.proposal.covmat)Adjust paths in agents/class-sz-engineer.md and the skill
allowed-tools if your venv lives elsewhere.
/class-sz:explain
How do I sample profile parameters at fixed cosmology with NUTS?
/class-sz:tszfast P0=10 beta=5.0
/class-sz:build-likelihood ACTClyy --data-dir "~/Desktop/class-sz-plugin-tests/data"
.claude-plugin/marketplace.json # single-plugin marketplace
plugins/class-sz/
.claude-plugin/plugin.json # plugin manifest
skills/
explain/SKILL.md # always-on classy_szlite knowledge
explain/reference.md # loaded on demand
tszfast/SKILL.md # /class-sz:tszfast
build-likelihood/SKILL.md # /class-sz:build-likelihood
agents/
class-sz-engineer.md # subagent
MIT
npx claudepluginhub borisbolliet/class-sz-claude-pluginSpecialized Claude assistance for classy_szlite — a pure-JAX cosmology code that covers CMB Cls, matter Pk (linear and nonlinear), distances, derived parameters, and the halo-model tSZ Cl^yy bandpower likelihood. Backed by the v2 CosmoPower emulators used in the ACT DR6 extended-cosmology analyses. Pipelines, gradient probes, NUTS/HMC sampling, and cobaya bandpower likelihood scaffolding.
Cobaya cosmological Bayesian-analysis framework assistant: scaffolding YAML inputs, theory/likelihood wiring (CAMB, CLASS, classy_szlite), MPI MCMC runs, getdist post-processing, debugging convergence (Rminus1) and acceptance issues.
Scientific paper writing in LaTeX (MNRAS, A&A, ApJ, JCAP). Scaffolds repos with the right class file + Makefile + bibtex layout, populates references via valency-mcp, builds + reports errors, bundles a knowledge skill with style rules and recipes learned from real submissions.
Google Cloud TPU v6e / v5e helper: create + bootstrap + connect + delete VMs via gcloud. Encodes the recipes and pitfalls from the tpu-2026 workspace (uv-not-PPA python3.12, libtpu for jax, flax install order, IAP-tunnelled SSH, Cloud NAT).
Specialized Claude assistance for the Cobaya cosmological Bayesian analysis framework.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive startup business analysis with market sizing (TAM/SAM/SOM), financial modeling, team planning, and strategic research
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Complete creative writing suite with 10 specialized agents covering the full writing process: research gathering, character development, story architecture, world-building, dialogue coaching, editing/review, outlining, content strategy, believability auditing, and prose style/voice analysis. Includes genre-specific guides, templates, and quality checklists.
Comprehensive .NET development skills for modern C#, ASP.NET, MAUI, Blazor, Aspire, EF Core, Native AOT, testing, security, performance optimization, CI/CD, and cloud-native applications