Guide accessibility reviews for UI, documents, and workflows. Use for WCAG-oriented audits, keyboard review, semantics, contrast, and remediation planning.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Qualify and assign a bounded local compact worker task through existing LSCli profiles and native tool calling, without adding a runner.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
설치 방법 보기설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.
Use this workflow to decide whether an existing local compact model/profile can safely complete one small, independent task through LSCli. The controller owns scope, grants, acceptance, and the checkpoint. Do not create another runner, silently change profile configuration, or treat model availability as task qualification.
Write a worker packet before dispatch: one observable outcome; exact read, disclosure, write, and recipe boundaries; allowed tools/files; checks; deadline/resource limits; checkpoint and acceptance criteria. Split coupled work before dispatch. If it cannot be split into an independently verifiable slice, keep it with the controller. Run at most two delegated assistants concurrently. For bounded research, implementation, and tests, follow the active client role configuration and choose the least costly available role capable of the task at an appropriate effort. Use an independent reviewer with capability matched to the risk for material-risk or policy-sensitive results. Preserve the user's active controller selection; do not change global routing.
Qualification has two separate gates: an OpenAI-compatible endpoint handshake proves only transport/schema compatibility; a successful observed native tool round trip through the actual LSCli runner is required before assigning tool-enabled work. Capture exact model ID, quantization, chat template, parser/tool-call format and revision. A local Gemma-family Q4 model is an example candidate, never a blanket-qualified model. Measure task completion, invalid tool calls, context use, latency, and resource use on representative bounded tasks before tuning. Preserve working profiles unless comparative evidence supports a change.
Follow the migration and smoke metadata in workflow.yaml. Canonical runner and permission contracts are in LSCLI.md and LSCLI_RUNTIME.md; compose LocalSetup Agent-* lane routing and task matching rather than duplicating them. The LocalSetup lane matrix recommends a task lane; it does not select a client model, effort, or worker configuration. Follow the active client's role configuration for those choices. The controller reviews the result against the packet and records the existing LSCli checkpoint; tool output or a worker report alone is not acceptance.
아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.
로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.
아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.
Guide accessibility reviews for UI, documents, and workflows. Use for WCAG-oriented audits, keyboard review, semantics, contrast, and remediation planning.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Use when choosing a LocalSetup Agent-* lane for exploration, research, bounded implementation, or review. Consult the static capability matrix; do not probe account or network availability.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Guide agent email inbox workflows with prompt-injection defenses. Use for inbound email processing, webhooks, routing, and safety review.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Navigate and understand codebases using agentlens hierarchical documentation. Use when exploring new projects, finding modules, locating symbols in large files, finding TODOs/warnings, or understanding code structure.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Use for Agent Q signed encrypted file-drop or mail ship and ingest, private registry v2 setup, queue status, and transport maintenance.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Inventory wrapper for the alirezarezvani Claude skills bundle. Use when triaging its broad skill catalog and large script surface before any targeted import.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.