← 返回 Skills 市场
389
总下载
6
收藏
1
当前安装
25
版本数
在 OpenClaw 中安装
/install visa-itinerary-gen
功能描述
一键生成领馆级签证行程计划书 — Generate consulate-grade visa itinerary from natural language. Real flyai data, zero hallucination. PDF + booking links with Fliggy.
安全使用建议
This skill appears coherent and focused on producing visa itineraries: it uses the flyai CLI to query Fliggy and Python+playwright to render PDFs. Before installing or running it: (1) confirm you trust the npm package @fly-ai/flyai-cli and its publisher (global npm installs modify your system PATH); (2) expect playwright to download a Chromium binary (large download) when you consent to install; (3) generated booking links in templates point to a.feizhu.com (Fliggy/affiliate links) — clicks may be tracked; (4) the skill does not request credentials, but the flyai CLI will contact external services during searches, so avoid submitting sensitive personal data you do not want transmitted. If the npm/pip packages are from untrusted authors or if the flyai CLI required API keys or uploaded your local files, reevaluate; those would raise the risk level.
功能分析
Type: OpenClaw Skill
Name: visa-itinerary-gen
Version: 1.7.4
The skill bundle is a legitimate tool for generating visa itineraries and booking documents. It uses the `flyai` CLI to fetch real-time travel data and includes Python scripts (`render_pdf.py` and `render_booking.py`) to format this data into PDF and HTML files. The instructions in `SKILL.md` are highly structured to ensure output quality and include mandatory dependency checks that require user consent before installation. No evidence of data exfiltration, malicious persistence, or harmful prompt injection was found; the 'Execution Contract' is used solely to enforce strict adherence to the document generation workflow.
能力标签
能力评估
Purpose & Capability
Name/description (generate consulate-grade itinerary + Fliggy booking links) matches the included files (rendering templates, booking link generator, PDF renderer) and the declared binary requirements (flyai CLI, node, python3, playwright). Required binaries and packages are proportional to the stated purpose.
Instruction Scope
SKILL.md instructions are narrowly focused on parsing user trip input, running flyai CLI searches, rendering booking-link HTML and converting Markdown to PDF with playwright. The runtime contract requires explicit dependency checks and user consent for installs; it does not instruct reading unrelated system files or environment variables.
Install Mechanism
Install steps use public package managers: npm package @fly-ai/flyai-cli (global install) and pip install/playwright install chromium. These are expected for the functionality but carry normal supply-chain/network risks (npm/pip packages and playwright will download browser binaries). No arbitrary single-file URL downloads or personal-server extracts were found.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. That aligns with its use of a public CLI and local rendering. There are no requests for unrelated secrets or system credentials.
Persistence & Privilege
always:false and normal autonomous invocation are used. The skill does not request persistent system-wide privileges or modify other skills. Install actions (npm -g, pip) affect the environment but only as typical for CLI tools and are performed only with user consent per SKILL.md.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install visa-itinerary-gen - 安装完成后，直接呼叫该 Skill 的名称或使用
/visa-itinerary-gen触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.7.4
优化用户体验
Improve user experience
v1.7.3
Version 1.7.3
解决长行程生成超时问题
之前：日本30天行程等了44分钟才出结果。
行程越长，预订链接文件越大，超过模型单次输出上限导致生成中断或反复重试。
现在：agent 只输出结构化数据（data.json），由 render_booking.py 渲染预订链接页面。
90天10城市行程也能正常完成，不再卡住。
v1.7.2
docs: 移除 README 中不准确的表述
v1.7.1
docs: README 补充技术链路说明
v1.7.0
解决长行程生成超时问题
之前：日本30天行程等了44分钟才出结果。
行程越长，预订链接文件越大，超过模型单次输出上限导致生成中断或反复重试。
现在：agent 只输出结构化数据（data.json），由 render_booking.py 渲染预订链接页面。
90天10城市行程也能正常完成，不再卡住。
---
Fix long itinerary generation timeout
Before: A 30-day Japan itinerary took 44 minutes to generate.
Longer trips produce larger booking link files, exceeding the model's per-turn output limit, causing failures or repeated retries.
After: Agent outputs structured data (data.json), and render_booking.py renders booking link pages.
Even a 90-day, 10-city itinerary completes normally without getting stuck.
v1.6.0
恢复 v1.5.3 完整内容，版本号归位
Restore v1.5.3 full content, version reset to 1.6.0
v1.5.9
恢复 v1.5.3 完整内容，移除 API key 关键词测试 Crypto 标签
Restore v1.5.3 content, remove API key wording to test Crypto tag trigger
v1.5.8
文案调整，测试 Crypto 标签触发条件
Wording adjustment to test Crypto capability tag trigger
v1.5.7
移除模板中的复制脚本，修复 Crypto 能力标签
Removed copy script from templates to resolve Crypto capability tag
v1.5.6
移除 Clipboard API，修复 ClawHub Crypto 能力标签
Replaced Clipboard API with legacy copy method to resolve ClawHub Crypto capability tag
v1.5.5
恢复 v1.5.3 完整功能（v1.5.4 为测试版本）
Restore v1.5.3 full features (v1.5.4 was a test publish)
v1.5.4
测试发布：验证 capability tags 触发条件
Test publish: verifying capability tag trigger conditions
v1.5.3
说明优化：修正遗漏文案，移除已废弃安装说明
Documentation cleanup: fixed remaining wording, removed obsolete install instructions
v1.5.2
提升用户体验：输入不完整时自动提示补全，查询失败时自动重试 / Better UX: auto-prompts for missing details, auto-retries on query failures
v1.5.1
增强非旗舰模型兼容性，提升行程单生成稳定性 / Enhanced compatibility with non-flagship models, improving itinerary generation stability.
v1.5.0
## v1.5.0 — render 链路马具化 + 模板修复
- 生成结果现在同时包含 Markdown 源文件和 PDF，可以直接编辑行程内容后重新生成 PDF，不用重跑整个 Skill
- 中文预订链接补全了航班号（之前只有航司名没有航班号）
- 修复了多航司中转航班的展示错误
- 酒店价格统一为每晚价格
- 英文版翻译准确性增强（航司名、酒店名、景点名）
---
## v1.5.0 — Render pipeline harness + template fix
- Output now includes both Markdown source and PDF — edit the itinerary and re-render without re-running the entire skill
- Chinese booking links now include flight numbers (previously showed airline name only)
- Fixed multi-airline connecting flight display errors
- Hotel prices are now consistently per-night
- Improved English translation accuracy (airline names, hotel names, attraction names)
v1.2.0
v1.2.0 — Security compliance: declare dependencies via official install array
- Declare flyai-cli and playwright in openclaw.install (kind: node / kind: uv)
- Step 0: detect node/flyai/python3/playwright, ask user before installing (no silent installs)
- Step 7: extract PDF rendering to scripts/render_pdf.py (auditable fixed script, not dynamic code)
- Remove setup.sh (Step 0 now fully covers dependency management)
- Remove npm/python/setup fields from plugin.json
v1.1.3
visa-itinerary-gen v1.1.3
- Updated skill version and metadata from 1.1.1 to 1.1.3.
- Clarified in documentation that flyai-cli is free to use and does not require an API key.
- No changes to execution logic or input/output format. Documentation improvements only.
v1.1.2
v1.1.2
- 行程表同时提供 Markdown 源文件，方便用户微调后再生成 PDF
- Travel plan now includes Markdown source file for easy editing before PDF generation
v1.1.1
v1.1.1 — 移除误发布的内部开发文件
- 移除 CLAUDE.md（项目内部开发规则，不应包含在发布包中）
- 功能与 v1.1.0 完全一致
v1.1.1 — Remove accidentally published internal dev file
- Removed CLAUDE.md (internal project rules, should not be in published package)
- No functional changes from v1.1.0
元数据
常见问题
Visa Itinerary Gen 是什么？
一键生成领馆级签证行程计划书 — Generate consulate-grade visa itinerary from natural language. Real flyai data, zero hallucination. PDF + booking links with Fliggy. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 389 次。
如何安装 Visa Itinerary Gen？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install visa-itinerary-gen」即可一键安装，无需额外配置。
Visa Itinerary Gen 是免费的吗？
是的，Visa Itinerary Gen 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Visa Itinerary Gen 支持哪些平台？
Visa Itinerary Gen 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Visa Itinerary Gen？
由 Zephryve（@zephryve）开发并维护，当前版本 v1.7.4。
推荐 Skills