By borisbolliet
Denario multi-agent research assistant over MCP: drive the end-to-end pipeline (setup -> idea -> methods -> results -> paper) through the denario MCP tools, configure per-role/multi-provider models and the VLM image-reviewer via params.yaml, and run/resume the cmbagent_lg analysis engine (plan/execute/restart).
How to drive the Denario research assistant over MCP — the end-to-end pipeline (setup → idea → methods → results → paper), configuring models/steps/VLM via params.yaml, and the cmbagent_lg analysis engine (plan/execute/restart, crash-recoverable). Use when the user wants to run Denario, call a denario_* or cmbagent_lg_* MCP tool, generate a research idea/methods/results/paper, or configure params.yaml.
Run the Denario research pipeline end-to-end over MCP — lay out a project from a data description, then idea → methods → results → paper, configured via params.yaml. Use when the user wants a full research run (not just one stage).
Specialized Claude Code assistance for driving the
Denario multi-agent research assistant over
MCP — the end-to-end pipeline (idea → methods → results → paper) and the
underlying cmbagent_lg analysis engine.
/denario:explain — passive knowledge skill: the two MCP servers (denario, cmbagent_lg), the pipeline stages and which tool runs each, the params.yaml control surface (per-role/multi-provider models, max_n_steps, the VLM image-reviewer), and crash-recoverable restart. Auto-loads when the conversation is about Denario, denario_results/denario_idea/…, cmbagent_lg, params.yaml analysis models, or restart_at_step./denario:run-pipeline <project_dir> [data description] — runs the full pipeline end-to-end through the denario MCP tools (lay out the project → idea → methods → results → paper), configured from params.yaml; reports a concise summary.This plugin tells Claude how to use the Denario MCP tools; you must have the two
servers registered (they ship with Denario, in denario/mcp_servers/):
LGPY=/Users/boris/pyvenvs/py312-cmbagent-lg/bin/python
DEN=/Users/boris/GitHub/Denario/denario/mcp_servers
claude mcp add denario -s user -- $LGPY $DEN/denario_server.py
claude mcp add cmbagent_lg -s user -- $LGPY $DEN/cmbagent_lg_server.py
See /denario:explain → reference.md for keys, the venv, and gotchas.
Local testing (no marketplace needed):
claude --plugin-dir ~/GitHub/denario-claude-plugin
From a marketplace (once published):
/plugin marketplace add borisbolliet/denario-claude-plugin
/plugin install denario@denario-claude-plugin
/denario:explain
How do I configure a multi-provider results run and turn on the image reviewer?
/denario:run-pipeline ~/Desktop/sho-study "A 1D damped harmonic oscillator: study amplitude/energy decay vs damping."
.claude-plugin/marketplace.json
plugins/denario/
.claude-plugin/plugin.json
skills/
explain/
SKILL.md # main knowledge skill (how to use the Denario MCP)
reference.md # tool signatures, full params.yaml schema, registration
run-pipeline/
SKILL.md # end-to-end pipeline over MCP
mcp__denario__*, mcp__cmbagent_lg__*). The run-pipeline skill pre-approves the pipeline tools in allowed-tools.denario_results is long and synchronous; for non-blocking runs use cmbagent_lg_execute + cmbagent_lg_status.denario_eda and the cmbagent-keyword paper path need the old cmbagent package, which is not in the cmbagent-lg venv — they're out of scope here.MIT
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.
npx claudepluginhub borisbolliet/denario-claude-plugin --plugin denarioSpecialized 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.
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.
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).
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Plugin-safe Claude Code distribution of Antigravity Awesome Skills with 1,561 supported skills.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer