Menu
Use when executing an implementation plan fully autonomously without human interaction
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Drive a native macOS app via the cua-driver CLI (default) or MCP server — snapshot its AX tree, click/type/scroll by element_index, verify via re-snapshot. Use when the user asks you to operate, drive, automate, or perform a GUI task in a real macOS application on the host (e.g. "open a file in TextEdit", "navigate to /Applications in Finder", "click the Save button in Numbers").
Firefox browser automation for AI agents using Playwright. Use when the user needs Firefox-specific browser automation, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or any task requiring Firefox. Triggers include requests to "use Firefox", "open in Firefox", "test in Firefox", "automate Firefox", or any browser task where Firefox is specified or preferred.
Use when the user wants browser automation against an existing Chrome or Chromium tab, especially to reuse a real logged-in session without reauthenticating. Triggers include requests to use the current browser, keep existing cookies or extensions, avoid login churn, or automate a live Chromium tab.
Remove AI writing patterns from prose. Use when drafting, editing, or reviewing text to eliminate predictable AI tells.
Use when validating UI changes end-to-end in any project before claiming work is complete — runs Playwright tests against the project's dev server to prove user stories work
Use when validating UI changes end-to-end before claiming work is complete — spawns a fresh server pair from the worktree and runs Playwright tests against it
| name | the-controller-autonomous-execution |
| description | Use when executing an implementation plan fully autonomously without human interaction |
Execute an implementation plan end-to-end with zero human interaction. Same quality gates as subagent-driven-development (fresh subagent per task, spec review, code quality review, final review), but the orchestrator handles all decisions autonomously.
Core principle: Full subagent-driven-development quality, zero human checkpoints — orchestrator resolves ambiguity from plan context, only surfaces when genuinely blocked.
Announce at start: "I'm using the the-controller-autonomous-execution skill to execute this plan autonomously. I'll report back when all tasks are complete."
digraph when_to_use {
"Have implementation plan?" [shape=diamond];
"Want fully autonomous execution?" [shape=diamond];
"the-controller-autonomous-execution" [shape=box];
"the-controller-subagent-driven-development" [shape=box];
"Use the-controller-writing-plans first" [shape=box];
"Have implementation plan?" -> "Want fully autonomous execution?" [label="yes"];
"Have implementation plan?" -> "Use the-controller-writing-plans first" [label="no"];
"Want fully autonomous execution?" -> "the-controller-autonomous-execution" [label="yes"];
"Want fully autonomous execution?" -> "the-controller-subagent-driven-development" [label="no - want human checkpoints"];
}
vs. subagent-driven-development:
vs. executing-plans:
digraph process {
rankdir=TB;
"Set up worktree (the-controller-using-git-worktrees)" [shape=box];
"Read plan, extract all tasks with full text, create TodoWrite" [shape=box];
subgraph cluster_per_task {
label="Per Task";
"Dispatch implementer subagent" [shape=box];
"Implementer asks questions?" [shape=diamond];
"Orchestrator answers from plan context" [shape=box];
"Orchestrator can answer?" [shape=diamond];
"Surface to human (ONLY blocker case)" [shape=box style=filled fillcolor=lightyellow];
"Implementer implements, tests, commits, self-reviews" [shape=box];
"Dispatch spec reviewer subagent" [shape=box];
"Spec compliant?" [shape=diamond];
"Implementer fixes spec gaps" [shape=box];
"Dispatch code quality reviewer subagent" [shape=box];
"Code quality approved?" [shape=diamond];
"Implementer fixes quality issues" [shape=box];
"Mark task complete" [shape=box];
}
"More tasks?" [shape=diamond];
"Dispatch final code reviewer for entire implementation" [shape=box];
"Report final summary" [shape=box style=filled fillcolor=lightgreen];
"Set up worktree (the-controller-using-git-worktrees)" -> "Read plan, extract all tasks with full text, create TodoWrite";
"Read plan, extract all tasks with full text, create TodoWrite" -> "Dispatch implementer subagent";
"Dispatch implementer subagent" -> "Implementer asks questions?";
"Implementer asks questions?" -> "Orchestrator answers from plan context" [label="yes"];
"Implementer asks questions?" -> "Implementer implements, tests, commits, self-reviews" [label="no"];
"Orchestrator answers from plan context" -> "Orchestrator can answer?";
"Orchestrator can answer?" -> "Dispatch implementer subagent" [label="yes - send answer"];
"Orchestrator can answer?" -> "Surface to human (ONLY blocker case)" [label="no"];
"Surface to human (ONLY blocker case)" -> "Dispatch implementer subagent" [label="human answers"];
"Implementer implements, tests, commits, self-reviews" -> "Dispatch spec reviewer subagent";
"Dispatch spec reviewer subagent" -> "Spec compliant?";
"Spec compliant?" -> "Implementer fixes spec gaps" [label="no"];
"Implementer fixes spec gaps" -> "Dispatch spec reviewer subagent" [label="re-review"];
"Spec compliant?" -> "Dispatch code quality reviewer subagent" [label="yes"];
"Dispatch code quality reviewer subagent" -> "Code quality approved?";
"Code quality approved?" -> "Implementer fixes quality issues" [label="no"];
"Implementer fixes quality issues" -> "Dispatch code quality reviewer subagent" [label="re-review"];
"Code quality approved?" -> "Mark task complete" [label="yes"];
"Mark task complete" -> "More tasks?";
"More tasks?" -> "Dispatch implementer subagent" [label="yes"];
"More tasks?" -> "Dispatch final code reviewer for entire implementation" [label="no"];
"Dispatch final code reviewer for entire implementation" -> "Report final summary";
}
Reuse from the-controller-subagent-driven-development:
the-controller-subagent-driven-development/implementer-prompt.mdthe-controller-subagent-driven-development/spec-reviewer-prompt.mdthe-controller-subagent-driven-development/code-quality-reviewer-prompt.mdWhen an implementer subagent asks a question, the orchestrator MUST attempt to answer it before surfacing to the human:
Only surface to human if: The answer genuinely cannot be determined from available context and guessing wrong would cause significant rework.
When answering: Dispatch a fresh implementer subagent with the original task text plus the answer. Provide the answer clearly, with reasoning, so the implementer can proceed confidently.
After all tasks complete and final review passes, report:
Never:
Always:
Required workflow skills:
Reuses prompts from:
Subagents should use:
Does NOT use: