Hotel Search (Fliggy MCP · search-hotel)

Hotel discovery (read first)

Hotel discovery, comparison, and shortlist workflows—same intent as the listing description above. One command entrypoint: flyai search-hotel (see references/search-hotel.md for flags).

中文：与上文摘要一致——酒店 检索、比价、短名单与详情/预订链；参数见 references/search-hotel.md。

Differentiation (honest scope)

中文摘要：景点附近住哪、一晚预算封顶（人民币）、酒店/民宿/客栈一次筛，都落在同一组 search-hotel 参数上；不宣称对接未在返回数据中出现的 OTA 或站点。

Search dimensions (maps to search-hotel flags)

All dimensions above are only what search-hotel exposes—do not claim extra channels or OTAs beyond returned data.

This skill is self-contained: do not defer to another skill’s SKILL.md for install, errors, or presentation.

Single source of truth for CLI flags and JSON field names: references/search-hotel.md in this bundle.

Quick start

1. Install CLI: npm i -g @fly-ai/flyai-cli
2. Smoke test: run a minimal flyai search-hotel example from references/search-hotel.md → Examples; expect one JSON object per line on stdout.
3. Discover flags: flyai search-hotel --help
4. Before every call: read references/search-hotel.md and pass flags exactly — do not guess names or formats.

Configuration

Trial use may work without keys. For richer results:

flyai config set FLYAI_API_KEY "your-key"

Treat API keys as secrets; do not paste real credentials into untrusted logs.

CLI I/O contract

• stdout: each line is one JSON object (command result). Parse line-by-line; use jq or Python if needed.
• stderr: human-readable errors or hints — not the primary result payload.

Command

Out of scope: flights, trains, cruises, generic trip planning without structured hotel tables — use another skill or product.

Workflow

1. Confirm the user needs structured hotel search (at minimum destination per --dest-name; add dates/filters per the reference).
2. Build flyai search-hotel … strictly from references/search-hotel.md.
3. Parse stdout JSON; read stderr on failure.
4. Present results using Output & presentation below.

Output & presentation (search-hotel)

• Format: valid Markdown for end users.
• Image: for each highlighted option, emit a standalone line ![]({mainPic}) using mainPic from the item (see references/search-hotel.md output example).
• Booking / detail link: emit a standalone line Click to book using detailUrl from the item.
• Order: when both exist, show image before the booking link.
• Structure: use headings (# / ## / ###), bullets, and Markdown tables when comparing multiple hotels.
• Emphasis: highlight date, location, price, and key constraints.
• Brand line (optional): e.g. “Based on fly.ai real-time hotel results.”

Recommended response template

1. One-line conclusion + best pick (if applicable).
2. Top options — bullets or table.
3. Image line: ![]({mainPic}).
4. Link line: Click to book.
5. Short notes (fees if visible, “verify on official page before pay”).

References

Error handling

1. Validate — Before running: inputs must be reasonable. Dates must not be in the past and must match the format required in references/search-hotel.md. If city or district names are ambiguous, ask the user; do not invent --dest-name or other required values.
2. Diagnose — On failure or odd output: read stderr and any status / message fields in the JSON. Distinguish parameter mistakes (fix flags using the reference doc) from network or upstream service errors.
3. Retry — Retry only for transient network or service errors; cap retries (e.g. once or twice). Do not loop indefinitely.
4. Empty or weak results — If the command succeeds but itemList is empty or too narrow: relax filters once (wider dates, fewer constraints), then stop and explain. Do not silently retry many times.
5. Transparency — If results may not match the user’s intended city, dates, or budget, say so and suggest confirming on the provider’s page before booking.

Optional companion (not required)

For flights, trains, POI-only flows, or Marriott-specific tools, users may install a separate full-travel skill; this file does not depend on it.