rudder

Automate Rudder control-plane issue workflows via CLI with JSON output.

280|30|Updated Mar 30, 2026
One-click install
npx skills add https://github.com/Undertone0809/rudder --skill rudder
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rudder
Source: https://github.com/Undertone0809/rudder/tree/main/server/resources/bundled-skills/rudder
Command: npx skills add https://github.com/Undertone0809/rudder --skill rudder

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It prevents agents from improvising Rudder control-plane actions by giving a reliable, CLI-first procedure for claiming work, reading the right context, and closing issues with correct governance signals.

Core Features & Use Cases

  • Heartbeat-ready orchestration: Run steps in the right order (identity → approval follow-up → inbox → mention wake → checkout → context → work → close-out).
  • CLI-first control plane: Use rudder ... --json as the stable interface, with explicit rules for when API fallback is acceptable.
  • Durable collaboration hygiene: Enforces checkout-before-work, structured reviewer decisions, and single close-out signals (comment/done/block/handoff).
  • Organization skill management: Supports scanning, importing, listing, and enabling org skills so agents can reuse operating patterns.
  • Planning and document handling: Updates issue documents with plan keys instead of appending to descriptions, and links back to #document-plan.

Quick Start

Use the rudder skill in your next heartbeat to claim one assigned issue, read its compact context, do the smallest useful chunk, and close it with exactly one structured close-out signal.

Frequently Asked Questions about rudder

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

FAQPage Schema
What is agent heartbeat orchestration for issue workflows?

Agent heartbeat orchestration sequences control-plane actions like identity loading, inbox prioritization, checkout, and close-out to turn issue assignments into safe, reviewable outcomes. It prevents agents from improvising actions during work management tasks.

How do I automate issue assignments using a CLI-first control plane?

You automate issue assignments by running CLI commands with the `--json` flag for deterministic discovery and mutations. This enforces a checkout-before-work procedure and structured close-out signals for reliable execution.

How do I enforce review governance for automated agent tasks?

You enforce review governance by applying structured close-out signals such as comment, done, block, or handoff. This ensures reviewer decisions are captured correctly and prevents unstructured or ambiguous issue closures.

When should I use API fallback instead of CLI for control-plane automation?

API fallback is acceptable only for diagnostic purposes or when encountering empty-stdout CLI defects. The CLI interface using `--json` remains the primary stable interface for all deterministic mutations and discovery.

What is the correct sequence for running a control-plane heartbeat?

The correct heartbeat sequence runs identity loading, approval follow-up, inbox prioritization, mention wake, checkout, context retrieval, work execution, and finally close-out. This order ensures proper governance and context before mutations.

How do I update issue planning documents without appending to descriptions?

You update issue planning documents by writing to `plan` keys within the document structure and linking back to `#document-plan`. This separates planning data from the main issue description for cleaner tracking.