City guide content generator
Produce a city guide from AGNT venue data with intent-based sections.
AGNT's venue data is structured but not editorial. This prompt transforms raw venue records into a city guide organized by intent patterns (date night, quick lunch, sunset drinks, co-working lunch) rather than cuisine type. The intent-based organization matches how users actually search.
The prompt reads venue data from the knowledge graph and produces markdown sections that can be published as a guide page or broken into individual signal posts.
The prompt
<<<
You produce a city guide from AGNT's venue data, organized by user intent.
INPUTS:
- city (e.g. "Canggu", "Seminyak", "Ubud")
- intent_patterns: list of intents to cover (e.g. ["date-night", "quick-lunch", "sunset-drinks", "coworking-lunch", "family-brunch"])
SEQUENCE:
1. Query venue knowledge graph for all venues in the city.
2. For each intent pattern, select 3-5 best-fit venues based on:
- Menu/cuisine match
- Atmosphere tags
- Price range
- Hours alignment (sunset drinks needs west-facing + open at 5pm)
3. Write a 2-3 sentence editorial blurb per venue (not a review — a recommendation).
4. Include: venue name, area, price range, best for, booking link.
OUTPUT per intent section:
## [Intent name]
[1-sentence section intro]
1. **[Venue name]** — [area] — [price range]
[2-3 sentence blurb]. [Booking link or "walk-in only"].
RULES:
- Only include venues with HIGH confidence data. No guessed hours or menus.
- Never include venues that haven't been verified in the last 90 days.
- If an intent pattern has < 3 qualifying venues, note it and skip.
- Write in AGNT brand voice: direct, local-knowledge, no fluff.
>>>When to use
Use this when producing or refreshing a city guide page. Run quarterly per city to keep recommendations current with venue data updates.
Works with
Related guides
Share as social post
City guide content generator (claude-code + mcp) — Produce a city guide from AGNT venue data with intent-based sections. https://agntdot.com/prompts/city-guide-generator
170 / 280 chars
Related prompts
Venue intake prompt pack for Claude Code
Parse a venue's website + menu into AGNT's knowledge pack format.
Ecosystem release-teardown prompt
The prompt the AGNT signals desk uses to draft ecosystem release breakdowns.
Venue reply drafting prompt
Draft a reply in the venue's voice, in the guest's language, with booking-aware context.