Convert ambiguous engineering requests into consensus-driven ADR-backed execution plans.
npx skills add https://github.com/xz1220/oh-my-kimi --skill ralplan-xz1220
Please help me install this Agent Skill. Skill: ralplan Source: https://github.com/xz1220/oh-my-kimi/tree/main/skills/ralplan Command: npx skills add https://github.com/xz1220/oh-my-kimi --skill ralplan-xz1220
Ralplan solves the problem of turning an underspecified task request into a safe, testable execution plan by forcing a structured consensus among Planner, Architect, and Critic before any implementation happens.
$plan --consensus flow to generate an ADR-backed plan instead of a single-shot outline.--deliberate flag that adds pre-mortem and expanded test planning.--interactive, it pauses at review/approval points to collect user decisions (proceed, request changes, reject) and then routes to execution (ralph/team/goal-mode) without directly implementing inside the planning skill.Use it when you need a robust plan for engineering work, especially when scope, acceptance criteria, or risk level is unclear, such as refactors, migrations, security-related changes, or production-impact tasks.
Run: $ralplan "task description"
High-intent search queries and answers about installing and using this skill.
Consensus-driven planning converts ambiguous engineering task requests into an ADR-backed execution plan by forcing structured agreement among Planner, Architect, and Critic iterations before any implementation happens.
The best way to plan high-risk engineering refactors with pre-mortem analysis is using a deliberate planning mode that adds pre-mortem evaluations and expanded test planning to safeguard risky production-impact tasks.
You add interactive approval gates to an agent workflow execution plan by using an interactive review flag that pauses the process at review points to collect user decisions like proceed, request changes, or reject before routing to execution.
Consensus planning supports sequential step ordering for complex migrations by converting underspecified scope and acceptance criteria into deliberate, testable execution steps without jumping straight into code implementation.
You should not use a single-shot outline when a task involves security-related changes or unclear risk levels, requiring deliberate-mode pre-mortem analysis and expanded test planning to safely establish engineering execution guardrails.
novel-jpRewrite weak task descriptions into structured instructions with explicit criteria.
BellaBeDerive goal states from execution data and generate corrective actions.
koasis89Coordinate Planner, Architect, and Critic agents for iterative consensus planning.
timbuchingerConvert vague requests into structured user stories with acceptance criteria or task lists.