steering-specs-generator

Generate steering documents and backlog items from guided engineering interviews.

Updated Apr 18, 2026
One-click install
npx skills add https://github.com/frznfrgg/codex_setup --skill steering-specs-generator-frznfrgg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: steering-specs-generator
Source: https://github.com/frznfrgg/codex_setup/tree/main/.agents/skills/steering-specs-generator
Command: npx skills add https://github.com/frznfrgg/codex_setup --skill steering-specs-generator-frznfrgg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Extract tacit engineering knowledge through guided interviews and generate structured steerings that guide agents to implement conventions and best practices.

Core Features & Use Cases

  • Conduct context-aware interviews to surface tacit knowledge and translate it into machine-readable steerings.
  • Generate standardized steering documents and action items from interview outcomes to align teams.
  • Use predefined packs or custom topics to capture domain-specific conventions and architectural rules.

Quick Start

Initiate an interactive interview flow to extract tacit engineering knowledge and generate steerings.

Frequently Asked Questions about steering-specs-generator

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

FAQPage Schema
How do I extract tacit knowledge from engineering teams into machine-readable conventions?

You capture tacit knowledge by running guided interviews that surface unwritten engineering rules, then translate outcomes into structured steering documents. This process codifies architecture invariants and project-specific norms into actionable formats for agents.

What are steering specs and when do I need them for my codebase?

Steering specs are structured documents codifying architecture invariants, coding conventions, and project-specific norms. You need them when unwritten engineering rules must be standardized into machine-readable formats to align teams and guide agents.

How do I document architecture invariants and coding conventions through interviews?

You document invariants by initiating an interactive interview flow that asks targeted questions about your codebase. Extracted tacit knowledge is then automatically generated into standardized steering documents and actionable backlog items.

Can I capture project-specific engineering norms for custom topics?

Yes, you can capture project-specific norms by applying the interview flow to custom topics rather than just predefined packs. This surfaces and codifies domain-specific conventions and architectural rules unique to your engineering environment.

What is the best way to generate actionable backlog items from engineering interviews?

The best way to generate backlog items from interviews is using a structured extraction process that produces both steering documents and corresponding session reports. These outputs translate tacit knowledge into actionable items for agents.

How do I align development teams on coding conventions and best practices?

You align teams by generating standardized steering documents from interview outcomes that codify best practices and coding conventions. These documents provide a shared, machine-readable reference point for all engineering members and agents.