What problem does it solve? Complex agent-driven engineering work often lacks structure: tasks get planned ad hoc, verification is skipped, subagents are misused, and PRs ship without independent review. This Skill imposes a disciplined operating style with named principles, task-matched playbooks, and mandatory verification so multi-step work stays auditable and correct. ## Core Features & Use Cases - Task-matched playbooks: Routes work to dedicated playbooks for bug fixes, features, refactors, investigations, perf hillclimbing, PR babysitting, shipping, and autonomous runs. - Principle-driven decisions: Applies named engineering principles (model the domain, prove it works, sequence verifiable units, guard the context window) with required citations to the decisions they changed. - Subagent orchestration: Defines defaults for background subagents, per-role model selection, swarm verification, and adversarial multi-model review before shipping. - Use Case: A user asks to fix a reported defect. The Skill routes to the Bug fix playbook: reproduce on the real surface, binary-search the root cause, delegate the fix to a subagent, verify the original repro passes, then open a PR with staged commits. ## Quick Start Ask the agent to work in poteto mode on your task, for example by saying "use poteto-mode to fix this bug and open a PR".
