| name | mother-goose |
| description | Route a user-described builder problem to exactly one bundled nursery-rhyme or story-based skill. Use only when the user explicitly invokes Mother Goose and asks for routing. Do not execute every story or silently activate during ordinary work. |
Mother Goose
Choose the tale whose lesson matches the problem. One problem receives one leading story.
Route
- Identify the repeated failure pattern, not merely the topic.
- Select exactly one:
- ring-around-the-rosie: remove one weakest option or unnecessary element per round;
- humpty-dumpty: stop repeating a repair strategy after its failure budget is exhausted;
- hickory-dickory-dock: checkpoint work and enforce a five-minute break at one hour;
- three-little-pigs: test resilience at straw, sticks, and bricks levels;
- goldilocks: calculate or justify the bounded middle setting;
- emperor-new-clothes: challenge groupthink, sycophancy, yes-man consensus, and authority bias by creating a protected evidence-based dissenting review;
- old-lady-swallowed-fly: stop cascading workarounds and trace the repair chain back to the earliest verified cause;
- sorcerers-apprentice: put hard scope, budget, stop, override, and cleanup controls around automation;
- pinocchio: require evidence for concrete build, test, compatibility, performance, and completion claims;
- hear-no-evil-speak-no-evil-see-no-evil: put an independent verifier under a strict read-only boundary so it can investigate and report defects to the Architect without inheriting Builder claims, inspecting future work, or performing repairs;
- boy-who-cried-wolf: separate real warnings from duplicate, unsupported, or false alarms without hiding fresh critical evidence;
- goose-golden-eggs: protect proven valuable behavior before optimizing or refactoring the component that produces it;
- five-little-monkeys: stop identical blind retries after the same failure signature exhausts its retry budget;
- hansel-and-gretel: leave recoverable operational breadcrumbs through a long or multi-agent build;
- tortoise-and-hare: make verified completion outrank raw speed when comparing work;
- little-red-hen: trace significant work, supplied evidence, and verification to the actor or source that actually performed it.
- Explain the lesson-to-problem match in one sentence.
- Do not execute the routed skill unless the user requested execution or confirms the route.
Output:
MOTHER GOOSE CHOOSES: [skill]
Lesson: [one sentence]
Why it fits: [one sentence]
If no story fits, say so. Never force a metaphor or invent a new skill during routing.
Completion
Finish when one justified route or a clear no-match result is delivered.
