Classify project type, recommend approach, identify needed agents, estimate complexity, and flag risks. Runs BEFORE /decompose to help North Starr adapt its pipeline to what is being built.
Installation
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Classify project type, recommend approach, identify needed agents, estimate complexity, and flag risks. Runs BEFORE /decompose to help North Starr adapt its pipeline to what is being built.
argument-hint
<requirement text, file path, or brief description>
Assess — Project Type Recommendation
Purpose
When a client gives a raw requirement, this skill classifies the project type and recommends the right approach BEFORE decomposition begins. It answers the questions that /decompose and chief-ai-po don't ask: what KIND of project is this, what agents will it need, and what's the right architecture approach?
Without /assess, the pipeline treats every project identically. With it, North Starr adapts its agent activation, pipeline configuration, and complexity expectations to the specific project type.
When to Use
Before /decompose — when a client gives a raw requirement and you need to understand what kind of project it is
Before /discover — when you want to validate your understanding of the project type before eliciting detailed requirements
On demand — when mid-project you realize the project type has shifted (e.g., what started as an automation pipeline now needs multi-agent orchestration)
For projects where the type is already clear and documented, skip directly to /decompose.
Input
The user provides one of:
Pasted text — a raw requirement, brief, or description
File path — a document (markdown, text, PDF) containing the requirement
Conversation summary — output from /discover or a prior discussion
Workflow
Step 1: Read & Understand the Requirement
Actions:
If a file path is provided, read the content (use PDF pagination for large PDFs)
If text is pasted, use it directly
Extract the core intent: what problem is being solved, for whom, at what scale
Identify any explicit constraints mentioned (budget, timeline, compliance, existing infrastructure)
Step 2: Classify Project Type
Evaluate the requirement against these project types:
Risk flags — check each. For every flag that applies, you MUST state (a) the concrete impact on this project and (b) a specific mitigation or next step. Do not list a risk without both.
Needs fine-tuning — domain-specific style or format that prompt engineering alone won't achieve
Needs existing infrastructure — depends on databases, APIs, or services not yet available
Needs credential provisioning — external API keys, OAuth flows, service accounts
High cost risk — large volume, expensive models, or open-ended generation
Novel problem — no established pattern in the team's experience
Multi-team dependency — requires coordination with other teams or external vendors
Example of a properly flagged risk:
Needs credential provisioning — Impact: Zendesk API requires OAuth2 service account; BUILD phase will block until credentials are provisioned. Mitigation: Request credentials in parallel with DESIGN phase; integration-planner will flag BLOCKED if missing at BUILD start.
