OpenParable
Skip to content
← Back to skills

Solution Intent

Documents
solution-intent

Solution Intent is an Agent Skill for documents published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Use when maintaining the solution intent document: scaffolding new projects (init), surgical updates after architectural changes (sync), or completeness validation (validate).

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/documents/solution-intent/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/solution-intent, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Solution Intent

Purpose

Manage the solution intent document (docs/solution-intent.md) lifecycle. The solution intent defines the architectural decisions, technical design, and evolution roadmap of the project. Three modes ensure the document stays accurate, complete, and fresh.

When to Use

  • New project needs a solution intent document -> init
  • Architectural changes were made (specs closed, stack changes, new agents/skills) -> sync
  • Pre-release or periodic health check -> validate
  • Automatically invoked by /ai-pr when staged changes include architecture files

Process

  1. Detect mode from arguments: init, sync, or validate
  2. Execute handler -- follow the matching handler in handlers/
  3. Report -- present summary of actions taken

Quick Reference

/ai-solution-intent init       # scaffold from template
/ai-solution-intent sync       # update sections from project state
/ai-solution-intent validate   # completeness and freshness scorecard

Integration

  • Called by: /ai-pr (step 6.7) when architectural changes detected
  • Calls: handlers/init.md, handlers/sync.md, handlers/validate.md
  • Reads: docs/solution-intent.md, .ai-engineering/manifest.yml, .ai-engineering/state/decision-store.json

Governance Notes

Visual priority: diagrams > tables > text. Every section MUST have at least one Mermaid diagram or table. Text accompanies but does not substitute visual representation.

TBD policy: if a section's data is not defined, implemented, or in scope, mark it explicitly as TBD. NEVER invent data.

Writing: use /ai-write patterns for document generation. The handler defines WHAT sections and data to gather; /ai-write defines HOW to write them.

Ownership: docs/solution-intent.md is project-managed. Sync updates data fields but never removes user-authored content. ai-eng update does not touch this file.

References

  • .agents/skills/pr/SKILL.md -- PR workflow that triggers sync automatically
  • .agents/skills/write/SKILL.md -- documentation writing patterns
  • .ai-engineering/manifest.yml -- governance surface counts, tooling, providers $ARGUMENTS