Skip to content

Prompt writing guide

PlanExe creates better plans when the input prompt is detailed and specific. Aim for 300–800 words.


What a good prompt includes

  • Goal and scope: what you want, and what you explicitly do not want.
  • Audience: who the plan is for (customers, users, stakeholders).
  • Constraints: budget, timeline, geography.
  • Location(s): country/region/city. If you have it, include the exact street address. Regulations and feasibility change by location.
  • Success criteria: what success looks like, and how it is measured.
  • Resources: team size, skills, existing assets, tools.

Good vs. weak prompts

Weak

Construct a bridge.

Better

Construct a bridge between Spain and Morocco across the Strait of Gibraltar. Target a feasibility‑to‑groundbreak timeline of 5 years with a total budget range of 8–12B EUR. The plan should cover geotechnical surveys, environmental impact assessments, maritime traffic coordination, and cross‑border permitting. Include options for rail + road, and describe staging for phase 1 (single rail) and phase 2 (road expansion). Exclude toll‑system design and focus on structural, logistics, and governance planning.


Goal:

Context / background:

Target users / customers:

Scope (in/out):

Constraints (budget, timeline, geography):

Location(s):

Success criteria:

Team / resources available:

Budget and money

Budget can be natural language. Examples:

  • A range: “$200k–$400k total.”
  • Phased: “$10M for phase 1, $5M for phase 2.”
  • With constraints: “Capex only, exclude staffing.”
  • Use standard currencies (EUR, DKK, RUB, BRL, etc.). Crypto budgets (BTC, ETH) are not supported.

Budget mistakes to avoid

  • Setting the budget to 0/none/N/A when the goal is serious and requires resources.
  • Using a currency code as the budget value (e.g., budget = DKK) instead of a real budget.

Common mistakes to avoid

  • Being too vague (one sentence).
  • Missing constraints (budget, time, scope).
  • Leaving out the location(s), which makes regulations and feasibility assumptions unreliable.
  • Conflicting requirements (e.g., “launch in 2 weeks” + “enterprise compliance”).