メニュー
SkillsMP は jaydubya818/MissionControl から 10 件の skill を収集しています。skill を開くとソースと詳細を確認できます。
収集済み skill 10 件中 10 件を表示しています。
Umbrella skill for integrating an agent with Mission Control, the orchestration and observability control plane. Use this skill when an agent needs to interact with Mission Control and you must route to the right focused skill — registration, heartbeats, task…
原文の言語: 英語
Use this skill when a Mission Control work item is complete or awaiting review and an operator needs a factual handoff draft with outcome, evidence, unresolved risk, next owner, and next action.
原文の言語: 英語
Tracks and enforces agent spend in Mission Control. Use this skill when recording costs via api.agents.recordSpend, declaring budgetDaily and budgetPerRun limits at registration, or reacting to budgetRemaining and budgetExceeded values returned by the…
原文の言語: 英語
Keeps an agent alive in Mission Control by calling the api.agents.heartbeat mutation every 30-60 seconds. Use this skill when running any agent session: agents that miss heartbeats for 2 minutes get quarantined, and the heartbeat response is how you discover…
原文の言語: 英語
Persists agent knowledge across sessions using the Mission Control agent documents system. Use this skill when depositing session memory before shutting down, keeping a WORKING_MD scratchpad up to date, writing daily notes, or reading back previously stored…
原文の言語: 英語
Registers an agent with Mission Control via the Convex api.agents.register mutation. Use this skill when an agent session starts and the agent must announce itself (name, emoji, role, workspace, task types, budgets) before doing any other work. Re-registering…
原文の言語: 英語
Requests human or dual-control approval for risky actions through the Convex api.approvals.request mutation. Use this skill when an action is YELLOW or RED risk tier — YELLOW needs a single approval, RED needs dual control (two decisions), and GREEN/LOW…
原文の言語: 英語
Records every agent execution turn as a run in Mission Control. Use this skill when starting a run (status RUNNING), completing or failing one (COMPLETED / FAILED), and reporting model, token counts, and cost so the observability and audit trail stays honest…
原文の言語: 英語
Submits finished work to Mission Control as content drops and posts progress to task threads. Use this skill when a piece of work is complete and needs human review, or when you should communicate progress, questions, or artifacts on a task — don't just…
原文の言語: 英語
Drives tasks through the Mission Control state machine: find INBOX tasks, assign yourself, and transition through ASSIGNED, IN_PROGRESS, REVIEW, and DONE. Use this skill when claiming, creating, or progressing a task — every status change must go through…
原文の言語: 英語