aops-working-disciplines

Selects and applies AOPS working disciplines for agent session workflows.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/eeemzs/aops --skill aops-working-disciplines-eeemzs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: aops-working-disciplines
Source: https://github.com/eeemzs/aops/tree/main/assets/skills/aops-working-disciplines
Command: npx skills add https://github.com/eeemzs/aops --skill aops-working-disciplines-eeemzs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Agents working on AOPS projects need a consistent way to choose and follow a working discipline—solo-pm-loop, build-review-chat, design-first-consensus, or coordinator-loop—so that review requests, consensus, memory, and closeout behavior stay aligned with the approved workflow instead of drifting ad hoc. ## Core Features & Use Cases - Discipline selection: Maps session signals (mode, agent count, uncertainty) to one of four canonical disciplines, with operator override via aops-cli start --discipline. - Mission policy seeding: Bridges aops-cli start --json output into hosted mission policy via aops-cli mission create/update --policy-json, including roles, guardrails, and signal mapping. - Recipes and guardrails: Provides per-discipline execution recipes (review-request per slice, consensus-to-plan binding gate, coordinator delegation rules) plus an explicit closeout checklist covering memory, review triage, issue triage, and resume readiness. - Use Case: An operator asks an agent to run a live implementer/reviewer session; the agent selects build-review-chat, creates a review request per slice with aops-cli pm review-request create, waits for accepted review results before committing with explicit pathspec, and closes out with a handoff memory entry. ## Quick Start Ask the agent to use the aops-working-disciplines skill to pick the right discipline for your session and seed the mission policy from aops-cli start output.

Frequently Asked Questions about aops-working-disciplines

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

FAQPage Schema
How do I choose between solo-pm-loop, build-review-chat, and design-first-consensus?▼

Choose based on uncertainty and session shape: solo-pm-loop for single-agent or low-uncertainty work, build-review-chat for live implementer/reviewer chat sessions, and design-first-consensus when design uncertainty is high or decisions are expensive to reverse. Use coordinator-loop when the operator delegates the session to one coordinator agent.

How do I seed mission policy from aops-cli start output?▼

Run aops-cli start --json and pass result.mission.policyJson into aops-cli mission create or mission update with the --policy-json and --apply flags. The policy convention includes discipline, guardrails, closeout, review, memory, and roles sections.

What is the consensus-to-plan binding gate in design-first-consensus?▼

It requires finalized consensus outputs with no _TBD_ placeholders, carrying the approved consensus ref into a PM task and sprint-backed implementation plan, and a reviewer accepting the plan approval review request before implementation begins. Operator approval is required before binding.

When should I use coordinator-loop instead of build-review-chat?▼

Use coordinator-loop when the operator wants a single coordinator agent as their only interface, delegating task definition, assignment, and review authority. Implementers route all questions to the coordinator and never contact the operator directly.

What does closeout require before leaving an AOPS session?▼

Closeout requires writing handoff memory with next action and refs, triaging open review requests and issues with owners, setting a truthful mission status, proving resume readiness, and recording a session summary. Each discipline adds specific items, such as accepted RRRs for build-review-chat.