[ Escoffier Labs · Academy ]
Deep Dives
Master-class teardowns of the systems behind the fleet, built to teach how each piece actually works. Every dive is authored as data and rendered by one engine, kept current with the tools it documents.
-
[ Brigade Master Class ]
Brigade, end to end
One canonical source for the MCP servers, tools, skills, and memory your AI coding agents share across 18 writer harnesses, merged into each tool’s native config with a review gate and a receipt for every change. A local operator-system CLI - not a runner, not a daemon, not a hosted service. Memory is one pillar of it, not the whole story. This is how every piece works, from the decide() memory guardrail to the MCP merge engine to the cross-model run loop.
brigade-cli v0.25.1Python 3.10+MITzero runtime deps -
[ Escoffier Labs Ecosystem ]
The whole kitchen
Brigade is the brigade de cuisine, but a kitchen is a team of stations. This is the Escoffier Labs fleet: which tools Brigade actually calls, the evidence pipeline that feeds it, and why the kitchen metaphor turns out to be the architecture.
Escoffier Labslocal-firstone tool per station -
[ Escoffier Labs Skill Roster ]
Skillet, skill by skill
A skillet full of production-tested agent skills: plain SKILL.md files that turn an idea into a shipped, reviewed change. 37 skills at v0.6.0, organized as a kitchen line: find work, design it, plan it, fire it, taste it, plate it, ship it.
skillet v0.6.037 skillsClaude Code pluginMIT -
[ Escoffier Labs · Course 11 (l’ardoise) ]
Mise en Scène: staging the explainer
A Vite + React 19 studio that turns raw source material into a self-contained, audience-tuned technical explainer rendered as a dark editorial SVG scene - exported as one standalone HTML file. An honest early spike, scoped without overclaiming.
v0.1.0 spikeReact 19 + Vitelocal studioMIT