docs: improve SEO for flow design reference pages

[ekline]

Context

This PR addresses SEO issues identified by auditing the Fastn documentation against the company's keyword strategy derived from fastn.ai.

Keyword map

Primary keywords: embedded integration platform, AI agent integrations, unified context layer, MCP gateway, connect AI agents to tools

Secondary keywords: workflow automation, managed OAuth, dynamic tool filtering, multi-tenant architecture, prebuilt connectors, AI workflow automation

Audience: SaaS companies, AI agent builders, platform engineers, developers building integrations

---

Top 3 SEO findings

Rank	Priority	Pages	Issue	Proposed fix
1	High	10 files in flows/flow-setup-essentials/designing-a-flow/	Missing meta descriptions — core product reference pages (connectors, loops, switches, variables, etc.) have no frontmatter, so search engines auto-generate poor snippets	Add description frontmatter with keyword-aligned summaries (120–160 chars)
2	High	~13 tutorial files in flows/tutorials/	Missing meta descriptions on tutorials — high-value SEO content with no description metadata for search result snippets	Add description frontmatter with action-oriented summaries
3	Medium	UCL pages in ucl-unified-context-layer/	Keyword misalignment — pages covering MCP functionality don't mention "MCP server" or "Model Context Protocol" in descriptions	Update descriptions to include MCP-related keywords

---

What this PR fixes

Finding #1: Added description frontmatter to all 10 flow design reference pages that were missing metadata entirely:

• connectors.md — prebuilt connectors, 250+ apps
• loop.md — loop components, workflow automation
• switch.md — conditional branching, automation paths
• variables.md — reusable values, flow configuration
• data-mapper.md — data transformation between steps
• database.md — SQL queries, structured data in flows
• converter.md — format conversion (JSON, CSV, XML, Parquet)
• custom-code.md — JavaScript, Python, C# custom logic
• download-file.md — file fetching for automation steps
• flow-response-success-and-error.md — HTTP response configuration

Each description is 120–160 characters, includes relevant keywords from the SEO strategy, and is action-oriented for better click-through rates in search results.

Remaining items

• Tutorial: Your first embedded integration widget #2: Add descriptions to ~13 tutorial pages in flows/tutorials/
• Tutorial: Build your first flow #3: Update UCL page descriptions to include MCP/Model Context Protocol keywords

How to verify

1. Check that each modified file has valid YAML frontmatter with a description field.
2. Preview the pages in GitBook to confirm descriptions render correctly as meta tags.
3. Verify no existing content was changed — only frontmatter was added.