Babysit a running multi-pane tmux agent sprint while the user is AFK or asleep. Polls the manager pane and grid persona panes every ~4 min via ScheduleWakeup, auto-approves routine agent permission prompts using a configurable rubric, escalates destructive prompts, and produces a morning summary. Use this skill whenever the user says "watch the sprint", "supervise overnight", "I'm going to bed keep the sprint going", "babysit the panes", "keep an eye on twingrid", or invokes "/sprint-supervisor" — even without the word "supervisor". Scales by named scope — one instance handles a twingrid (18 panes), multiple instances divide-and-conquer a triplegrid or quadgrid by scoping each supervisor to a slice of grids coordinated via /tmp lockfile so peers don't double-approve. Pairs with sprint-watchdog.sh which absorbs the mechanical ~70% of approvals; this skill handles the judgment ~30% — commands, rate-limit prompts, escalation.
Install with
Claude CodeCodexCursorGitHub CopilotGemini CLI
Supported by the Atlas install prompt. Skill-specific requirements vary.
Skill stats
Skills1CategoryProduct & designStars0
About this skill
Babysit a running multi-pane tmux agent sprint while the user is AFK or asleep. Polls the manager pane and grid persona panes every ~4 min via ScheduleWakeup, auto-approves routine agent permission prompts using a configurable rubric, escalates destructive prompts, and produces a morning summary. Use this skill whenever the user says "watch the sprint", "supervise overnight", "I'm going to bed keep the sprint going", "babysit the panes", "keep an eye on twingrid", or invokes "/sprint-supervisor" — even without the word "supervisor". Scales by named scope — one instance handles a twingrid (18 panes), multiple instances divide-and-conquer a triplegrid or quadgrid by scoping each supervisor to a slice of grids coordinated via /tmp lockfile so peers don't double-approve. Pairs with sprint-watchdog.sh which absorbs the mechanical ~70% of approvals; this skill handles the judgment ~30% — commands, rate-limit prompts, escalation.
Quick start
Copy the install prompt. Use the green button above this README.
Send it to your agent. Paste it into your coding agent’s chat and send.
Review the setup. Your agent checks the skill’s instructions, requirements, and installation scope.
Babysit a running multi-pane tmux agent sprint while the user is AFK or asleep. Polls the manager pane and grid persona panes every ~4 min via ScheduleWakeup, auto-approves routine agent permission prompts using a configurable rubric, escalates destructive prompts, and produces a morning summary. Use this skill whenever the user says "watch the sprint", "supervise overnight", "I'm going to bed keep the sprint going", "babysit the panes", "keep an eye on twingrid", or invokes "/sprint-supervisor" — even without the word "supervisor". Scales by named scope — one instance handles a twingrid (18 panes), multiple instances divide-and-conquer a triplegrid or quadgrid by scoping each supervisor to a slice of grids coordinated via /tmp lockfile so peers don't double-approve. Pairs with sprint-watchdog.sh which absorbs the mechanical ~70% of approvals; this skill handles the judgment ~30% — commands, rate-limit prompts, escalation.
About this listing
Category
Product & design
Reported license
MIT
Listing language (estimated)
English
Review status
Not individually reviewed
Agent compatibility
Atlas installation supports Claude Code, Codex, and Cursor. Skill-specific dependencies and features are checked during setup.