generate-feature-expectation

Generate expectation.yaml with success_scenarios and recovery entries from feature intent.

3|Updated Jan 25, 2026
One-click install
npx skills add https://github.com/kapilvirenahuja/garura --skill generate-feature-expectation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: generate-feature-expectation
Source: https://github.com/kapilvirenahuja/garura/tree/main/core/components/skills/generate-feature-expectation
Command: npx skills add https://github.com/kapilvirenahuja/garura --skill generate-feature-expectation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns a feature’s intent triple plus its available context into a testable, recoverable expectation artifact so teams can verify behavior deterministically instead of relying on ambiguous descriptions.

Core Features & Use Cases

  • Success-scenario generation: Produces vetted.pending success_scenarios as observable, binary measures for each actor-outcome pair.
  • Recovery generation: Produces exactly one recovery entry per failure condition with runtime triggers, directional guidance, and handoff rules (autonomous vs human).
  • Context-aware grounding: Reads context when provided to ground runtime personas and measures, and clearly downgrades to low grounding when context is missing.

Quick Start

Use generate-feature-expectation to write expectation.yaml from your feature intent.yaml and an optional Context bundle directory.

Frequently Asked Questions about generate-feature-expectation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate testable feature expectations from an intent document?

Feature expectations are generated by processing a feature intent triple and optional context bundle to produce an expectation.yaml artifact containing success scenarios and recovery entries. This grounds behavior in deterministic, observable measures rather than ambiguous descriptions.

What is runtime expectation generation in intent-driven development?

Runtime expectation generation creates testable artifacts at execution time within the craft-ice flow. It reads rules and intent paths to output success scenarios with binary measures and recovery entries with runtime triggers for failure conditions.

How do I create recovery scenarios for feature failure conditions?

Recovery scenarios are created by generating exactly one recovery entry per failure condition with runtime triggers, directional guidance, and handoff rules. The process distinguishes between autonomous and human handoffs within the expectation artifact.

Does expectation generation work without a context bundle?

Expectation generation works without a context bundle by downgrading to low grounding. When context is missing, runtime personas and measures are generated without context-aware grounding, though the resulting expectation.yaml still stamps vetted status as pending.

How do I structure success scenarios for actor-outcome pairs?

Success scenarios are structured as vetted pending observable, binary measures for each actor-outcome pair. The generation process reads intent from specified paths and writes these measures to the output path as part of the expectation layer.

What limitations exist when generating feature expectations at runtime?

Limitations include requiring valid intent triples and reading rules from specified paths, restricting writes to a single output_path, and downgrading context grounding quality when context bundles are missing or incomplete during craft-ice flow execution.