
buccaneermethodology/dashboard-governance-skill
Developer toolsこのページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る
Codex skill for maintaining Big Ideas, Sessions, Decisions, and emergent next steps in a project Dashboard
概要
Current candidate version: 0.4.1. Local files and validators do not prove publication; publication requires the tag and GitHub Release steps in docs/PUBLISHING.md. dashboard-governance is a portable Codex skill for maintaining a project Dashboard across long-running human-AI work. It helps an AI coding agent update Big Ideas, Sessions, Decisions, TSP fields, status, blockers, durable outputs, scope-aware closeout, and emergent next steps without treating the Dashboard as canonical project truth. The skill is useful when a repository needs an execution-state layer that survives model changes, session changes, and handoff between humans and AI agents. The default template stays lightweight; v0.2.0 also adds an optional advanced governance layer for projects that need stage plans, risks, exceptions, quality metrics, automation, external artifact indexes, agent logs, or validation/closure lanes.
README
以下は原文 README の抜粋です。原文を見る。
Dashboard Governance Skill
Current candidate version: 0.4.1. Local files and validators do not prove publication; publication requires the tag and GitHub Release steps in docs/PUBLISHING.md.
dashboard-governance is a portable Codex skill for maintaining a project Dashboard across long-running human-AI work. It helps an AI coding agent update Big Ideas, Sessions, Decisions, TSP fields, status, blockers, durable outputs, scope-aware closeout, and emergent next steps without treating the Dashboard as canonical project truth.
The skill is useful when a repository needs an execution-state layer that survives model changes, session changes, and handoff between humans and AI agents. The default template stays lightweight; v0.2.0 also adds an optional advanced governance layer for projects that need stage plans, risks, exceptions, quality metrics, automation, external artifact indexes, agent logs, or validation/closure lanes.
What It Is For
Use this skill when you want an AI agent to:
- keep project execution state visible across sessions
- separate stable truth from active task tracking
- maintain a Dashboard with Big Ideas, Sessions, and Decisions
- preserve Topic, Scope, and Purpose for large tracks and bounded tasks
- update task state after meaningful progress
- record unresolved choices separately from implementation tasks
- close narrowed work honestly before marking rows done
- identify concrete P0/P1 next-session candidates after work finishes
- preserve P2/P3 candidate memory only when it is specific and costly to rediscover
- use optional advanced governance files for non-trivial governed batches
- keep
todorows as visible options, not execution promises
It is not a replacement for an architecture document, issue tracker, design review, or human priority decision. It is a lightweight coordination surface for agentic work.
Core Concepts
Big Ideas
Big Ideas are long-lived semantic tracks. They describe important directions, themes, or implementation streams that may span many work sessions. A Big Idea should not close just because one related session finishes.
Sessions
Sessions are bounded execution or exploration units. They are small enough to be selected, completed, reviewed, and handed off. Sessions carry priority, status, dependencies, deliverables, exit criteria, next steps, and notes.
Decisions
Decisions capture choices that could block or distort multiple sessions if left implicit. Keeping them separate from Sessions prevents unresolved strategy from becoming hidden task churn.
TSP
推奨ツール
別のキーワードを試すか、フィルタを外してください。
インストール
npx skillfish add buccaneermethodology/dashboard-governance-skill