Prompt patterns we actually drill
- Skeleton + guardrail blocks for product briefs.
- Pairwise comparison scripts instead of vanity scores.
- Routing tables with explicit fallbacks when vendors shift tokenizers.
- Diff-first prompts for coding assistants with rollback language.
- Notebook contracts that survive audits without slowing researchers.
Use-case categories on the floor
Operations groups
Ticket triage, routing playbooks, incident comms prompts with human checkpoints.
Product + design trios
Critique prompts with principle tags, decision logs, and accessibility constraint tables.
Analyst desks
Evidence tables, memo scaffolding, and retrieval hints that respect sourcing policy.
Technical learners
Automation hooks, smoke prompts after deploy, notebook hygiene for exploratory prompting.
Mentor-led labs (what changes in the room)
Labs are not lectures. Facilitators cold-read your artifacts, mark ambiguity where prompts will fail, and run pairwise or diff exercises live. You leave with red ink you can defend to leadership.
- Live pairwise calibration with a second facilitator cross-checking scores.
- CI desk hours for automation tracks—bring your YAML, leave with a failing test you are proud of.
- Async video critiques for design tracks—decisions stay attached to files, not chat drift.
Learner FAQs
This page sequences the same ideas as a blueprint: patterns first, then categories, then labs. The catalog page is for filtering and comparison.
Signal flow diagram
Artifacts in
|
v
Pattern forge (skeletons / grids / routing)
|
v
Calibration or CI gates
|
v
Markdown export for leadership
|
v
Retro + archive list
Contact (exact studio coordinates)
4000 Alameda Macedo Marli de Nossa Senhora, Acre 55054-732 Brazil
Open the inbox form