スキル一覧に戻る
abdelrahman146

lane-playbooks

by abdelrahman146

Kyora is a SAAS service that helps social media store owners manage their business and finances

0🍴 0📅 2026年1月24日
GitHubで見るManusで実行

SKILL.md


name: lane-playbooks description: 'Kyora Agent OS lane workflows: Discovery, Planning, Implementation, Review, Validation, Recovery. Use when routing tasks through lanes, understanding lane-specific outputs, applying lane handoffs, or resuming work in a new session. Triggers: lane, discovery, planning, implementation, review, validation, recovery, handoff, continue.'

Lane Playbooks

Operational playbooks for each lane in the Kyora Agent OS continuous cycle. Each lane has specific entry conditions, outputs, DoD, and token-efficient approaches.

When to Use This Skill

  • Routing a task to the correct lane
  • Understanding what a lane should produce
  • Handling lane transitions and handoffs
  • Resuming unfinished work (Recovery lane)
  • Ensuring correct lane DoD before handoff

Prerequisites

  • Familiarity with KYORA_AGENT_OS.md operating model
  • Access to workspace for validation commands (Implementation/Validation lanes)

Lane Overview

LaneEntry ConditionPrimary OwnerNext Lane Default
DiscoveryUnclear bug; unknown area; cross-stackOrchestrator / LeadPlanning or Deferred
PlanningMedium/high risk; cross-stack; UX changeLeadImplementation
ImplementationPlan approved or low-risk clear changeImplementerReview
ReviewBefore declaring "done"Lead / QAValidation
ValidationAfter changes landQA / ImplementerRelease/Follow-up
RecoveryNew session continues unfinished workOrchestratorPrevious lane

Lane Quick Reference

Discovery

  • Entry: unclear bug; unknown area; cross-stack unknown
  • Output: findings + hypotheses + repro steps + next plan
  • DoD: can explain the smallest next step with high confidence
  • Token playbook: search first; read only relevant slices

Planning

  • Entry: medium/high risk; cross-stack; large refactor; UX changes
  • Output: phased plan + acceptance checks + gates
  • DoD: phases are verifiable; contracts and boundaries explicit
  • Token playbook: keep plan compact; link SSOT files

Implementation

  • Entry: plan approved or low-risk change
  • Output: code changes + tests; minimal notes
  • DoD: acceptance criteria met; touched-area checks pass
  • Token playbook: small change → validate immediately

Review

  • Entry: before declaring "done"
  • Output: checklist results + fix list
  • DoD: gates passed; consistency verified

Validation

  • Entry: after changes land
  • Output: command evidence (tests/build/e2e) + any follow-ups
  • DoD: relevant checks green; unrelated failures triaged

Recovery

  • Entry: new session continues an unfinished task
  • Output: Recovery/Resume Packet (required before continuing)
  • DoD: objective reconstructed; next smallest step identified

References

Handoff Rules (All Lanes)

Handoffs are mandatory when:

  1. Switching lane owners (e.g., Lead → Implementer)
  2. Switching scope (single-app → cross-stack)
  3. Pausing mid-task (token/time)
  4. Starting a new session for unfinished work

Minimum rule: no role starts Implementation without a task packet + a handoff packet (unless tiny, low-risk change).

Use the agent-workflows skill for all packet creation:

  • Delegation Packet: Lead → Implementer (Workflow 2 in agent-workflows)
  • Phase Handoff Packet: End of a phase (Workflow 3 in agent-workflows)
  • Recovery Packet: New session continues unfinished work (Workflow 4 in agent-workflows)

See: agent-workflows/SKILL.md

Note: The deprecated prompts (/create-delegation-packet, /create-phase-handoff-packet, /create-recovery-packet) are PO-only and cannot be triggered by agents.

Success Checklist

  • Correct lane selected based on risk + scope
  • Lane entry conditions verified
  • Lane-specific outputs produced
  • DoD for lane met before handoff
  • Handoff packet created when required
  • No "surprise docs" generated
  • Stop-and-ask triggers checked (see quality-gates)

Stop-and-Ask Triggers

MUST ask PO before proceeding if any are true:

  • Acceptance criteria missing and behavior ambiguous
  • Schema changes or migrations needed
  • Auth/RBAC/tenant boundary touched
  • Breaking API contract implied
  • New dependency needed

Troubleshooting

IssueCauseSolution
Skill not activatedDescription keywords don't matchUse explicit trigger words (lane, discovery, planning, etc.)
Wrong lane selectedRisk/scope misclassifiedRe-evaluate using examples.md quick reference
Handoff packet missing contextRushed handoffUse handoff prompts, verify all fields completed
DoD unclear for laneLane reference not consultedRead specific lane reference file

Validation Commands

Discovery/Planning lanes have no executable commands (output is plan/packet).

For Implementation/Validation lanes, run repo validation:

# Backend changes
make test.quick        # Unit tests
make openapi.check     # OpenAPI alignment

# Portal changes
make portal.check      # Lint + typecheck

# Full validation
make test              # All backend tests
make portal.build      # Portal build

SSOT References

Do not duplicate these rules; link to them:

スコア

総合スコア

50/100

リポジトリの品質指標に基づく評価

SKILL.md

SKILL.mdファイルが含まれている

+20
LICENSE

ライセンスが設定されている

+10
説明文

100文字以上の説明がある

0/10
人気

GitHub Stars 100以上

0/15
最近の活動

3ヶ月以内に更新がある

0/10
フォーク

10回以上フォークされている

0/5
Issue管理

オープンIssueが50未満

+5
言語

プログラミング言語が設定されている

+5
タグ

1つ以上のタグが設定されている

0/5

レビュー

💬

レビュー機能は近日公開予定です