メニュー
Enhance a hook-scaffolded EJS session journey file with semantic content (problem/intent, agents involved) and confirm initialization.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Capture sub-agent decisions, rationale, and inter-agent handoffs during multi-agent Engineering Journey System (EJS) workflows. Ensures that delegated work is properly attributed and traceable.
Analyze a Product Requirements Document (PRD) and generate a complete team of GitHub Copilot custom agents and reusable skills tailored to the project. Use this skill when asked to create, scaffold, or design a development team from a PRD or technical specification.
Build a Feature PRD that captures a new feature request in the context of an existing, completed project. Use this skill when the initial PRD has been implemented and you need to add a new feature without regenerating the entire project plan or agent team.
Build a comprehensive Product Requirements Document (PRD) or Technical Specification from a user's idea, concept, or research document. Use this skill when asked to create, draft, or formalize a PRD, spec, or requirements document.
Initialize and manage EJS artifacts throughout session lifecycle (Session Journey initialization, continuous updates, and finalization; ADR only when significant)
| name | ejs-session-init |
| description | Enhance a hook-scaffolded EJS session journey file with semantic content (problem/intent, agents involved) and confirm initialization. |
Use this skill when a session is starting — for example when the user says "initialize session", "start session", "create session journey", or begins work on a new task, feature, or bug fix.
Note: Copilot hooks (
scripts/hooks/session-start.sh) automatically create the journey file scaffold and sync the EJS database at session start. This skill enhances the scaffold with semantic content that hooks cannot provide.
Locate the journey file
ejs-docs/journey/YYYY/ejs-session-YYYY-MM-DD-<seq>.mdejs-docs/journey/_templates/journey-template.mdPopulate semantic metadata Fill in the fields that hooks cannot determine:
author — the human user (if known)agents_involved — list the active agents (e.g., [copilot])Capture initial Problem / Intent
Confirm initialization
"Session initialized: ejs-session-YYYY-MM-DD-<seq>"ejs-docs/journey/_templates/journey-template.mdejs-docs/adr/0000-adr-template.mdejs-docs/session-lifecycle-patterns.mdscripts/adr-db.pyscripts/hooks/session-start.shInitialize early, capture context while it's fresh. Hooks guarantee the structural scaffold; this skill adds the semantic meaning that only an LLM can provide.