agpair
Overview
Use this skill when your AI coding agent is the reviewer/controller and Antigravity is the executor.
INLINECODE0 is the control surface for:
- - preflight health checks
- task dispatch
- terminal-phase waiting
- evidence review
- semantic follow-up (
continue, approve, reject, retry)
It is not a second orchestrator and it is not the semantic decision-maker.
Triggering
This skill is intended to trigger when the user asks their AI agent to:
- - send or delegate work to Antigravity
- use INLINECODE5
- inspect
doctor, daemon, task status, or INLINECODE9 - review an INLINECODE10
- approve, reject, continue, or retry a delegated task
For the strongest activation, the user can explicitly say use agpair or send this to Antigravity via agpair.
Workflow
1. Preflight first
Before any semantic action, check:
- - INLINECODE13
- INLINECODE14
Do not continue if the target repo is unhealthy:
- - INLINECODE15
- INLINECODE16
2. Inspect task truth
Use:
- - INLINECODE17
- INLINECODE18
Do not choose continue, approve, reject, or retry until status and logs were read.
3. Blocking wait discipline
If you enter a blocking wait path (task start default --wait, task wait, or semantic commands with default --wait):
- - treat the wait as an active operation until it exits
- keep consuming the same long-running command session, or keep checking:
-
agpair task active-waits
-
agpair task status <TASK_ID>
- - do not tell the user the task is done while an active waiter still exists
INLINECODE29 means accepted, not completed.
4. Guard against premature intervention
If agpair task active-waits shows the task, or task status shows waiter_state=waiting:
- - do not send another semantic action on the same task
- do not abandon/retry the task
- only use
--force if the waiter is clearly orphaned
5. Pick one semantic action
Choose exactly one:
- -
continue for same-session follow-up - INLINECODE35 when evidence is good enough for finalization
- INLINECODE36 when work must continue in the same session
- INLINECODE37 only when the session is stale or not worth continuing
Required gates
Before claiming completion:
- - health was checked
- current task status was checked
- latest logs were checked
- if blocking wait was started, polling continued until terminal exit
- no same-task semantic action was sent while an active waiter existed
Anti-patterns
- - Do not start a blocking wait and then stop polling while the conversation is still alive.
- Do not treat
ACK as proof of progress. - Do not jump straight to
continue because the user said “继续”. - Do not hide
desktop_reader_conflict or repo_bridge_session_ready=false. - Do not invent commands or transport paths outside the real
agpair CLI.
agpair
概述
当你的AI编码代理担任审核/控制角色,而Antigravity作为执行者时,使用此技能。
agpair是以下功能的控制面板:
- - 飞行前健康检查
- 任务分发
- 终端阶段等待
- 证据审核
- 语义后续操作(continue、approve、reject、retry)
它既不是第二个编排器,也不是语义决策者。
触发条件
此技能旨在当用户要求其AI代理执行以下操作时触发:
- - 向Antigravity发送或委派工作
- 使用agpair
- 检查doctor、daemon、task status或task logs
- 审核EVIDENCE_PACK
- 批准、拒绝、继续或重试已委派的任务
为获得最强激活效果,用户可以明确说出use agpair或send this to Antigravity via agpair。
工作流程
1. 先做飞行前检查
在执行任何语义操作之前,检查:
- - agpair doctor --repo-path <绝对仓库路径>
- agpair daemon status
如果目标仓库不健康,请勿继续:
- - desktopreaderconflict=true
- repobridgesession_ready=false
2. 检查任务真实状态
使用:
- - agpair task status ID>
- agpair task logs ID> --limit 20
在读取状态和日志之前,不要选择continue、approve、reject或retry。
3. 阻塞等待纪律
如果进入阻塞等待路径(task start默认--wait、task wait或默认带--wait的语义命令):
- - 将等待视为持续操作,直到其退出
- 持续消费同一长时间运行的命令会话,或持续检查:
- agpair task active-waits
- agpair task status
- - 当活跃等待者仍然存在时,不要告诉用户任务已完成
ACK表示已接受,而非已完成。
4. 防止过早干预
如果agpair task active-waits显示该任务,或task status显示waiter_state=waiting:
- - 不要对同一任务发送另一个语义操作
- 不要放弃/重试该任务
- 仅在等待者明显孤立时使用--force
5. 选择一个语义操作
精确选择其中一个:
- - continue用于同会话的后续操作
- approve当证据足够好以完成最终确定
- reject当工作必须在同一会话中继续
- retry仅当会话已过时或不值得继续时
必需的门控条件
在声称完成之前:
- - 健康检查已完成
- 当前任务状态已检查
- 最新日志已检查
- 如果启动了阻塞等待,持续轮询直到终端退出
- 在活跃等待者存在时,未发送同一任务的语义操作
反模式
- - 不要启动阻塞等待后,在对话仍活跃时停止轮询。
- 不要将ACK视为进展的证据。
- 不要因为用户说继续就直接跳到continue。
- 不要隐藏desktopreaderconflict或repobridgesession_ready=false。
- 不要在实际的agpair CLI之外发明命令或传输路径。