Project Orchestrator

Problem It Solves

Many projects get stuck between idea and execution: the input may be one sentence, a screenshot, or a shared link, but the work still requires defining goals, choosing a plan, allocating steps, tracking progress, and delivering a result. Project Orchestrator positions itself as an AI COO, not just a command-runner. The user provides direction and minimal decisions, while the skill translates vague requests into an executable project and manages the process and delivery.

How It Works

• Idea translation: accepts text, voice, images, links, or files, then extracts goals and asks up to 3 clarifying questions.
• Plan design: generates 2–3 options, such as fast / balanced / premium, with a recommendation. Listed archetypes include content production, tooling, data engineering, deep research, creative design, and system integration.
• Lightweight decisions: users can confirm with ok, switch, urgent, slow, stop, or an emoji; complex decisions are broken into staged checks.
• Quiet execution: routine issues are handled automatically, while spending, risk, or failure triggers pause and confirmation.
• Delivery shape: outputs usable results, core conclusions, and reusable methodology.

Boundaries

It is best suited for project-style work that needs planning, decomposition, multi-step execution, and acceptance. Sensitive operations, external spending, or high uncertainty require explicit confirmation or clear limitation. The skill does not handle illegal or harmful tasks. At startup, define acceptance criteria, budget limits, and risk thresholds to keep quiet execution within authorized scope.