Menu
Refine rough ideas into concrete proposals through divergent and convergent thinking. Use when a concept needs product or technical shaping.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | idea-refine |
| description | Refine rough ideas into concrete proposals through divergent and convergent thinking. Use when a concept needs product or technical shaping. |
| license | MIT |
| compatibility | Works with coding agents that support Agent Skills. Upstream source is Addy Osmani's agent-skills repository. |
| metadata | {"agentpacks.version":"0.1.0","agentpacks.source":"https://github.com/addyosmani/agent-skills/tree/main/skills/idea-refine","agentpacks.upstream":"addyosmani/agent-skills/idea-refine"} |
This registry entry references the upstream idea-refine Agent Skill. Installers should use agentpacks.source and avoid copying upstream content into this registry.
Create, critique, and operationalize engineering strategy documents using diagnosis, tradeoff exploration, guiding policy, coherent actions, and operating cadence. Use when an engineering leader needs strategy for a team, platform, architecture, migration, reliability program, or multi-quarter technical investment.
Run engineering leadership reviews across strategy, delivery, quality, architecture, launch readiness, security risk, and performance health.
Inspect GitHub pull requests for leadership-relevant release risk, quality gates, ownership, and follow-up actions.
Keep dependencies secure, minimal, and up to date. Use when auditing a project's supply chain, adding new libraries, evaluating upgrades, or hardening a CI pipeline against dependency risks.
Design and implement GraphQL APIs with correct schema design, resolver patterns, N+1 prevention, federation, and security. Use when building or evolving a GraphQL layer.
Make ML experiments reproducible and comparable by tracking hyperparameters, metrics, artifacts, and environments. Use when training models, tuning hyperparameters, or promoting models to production.
