feat: add Langdock agent completions docs

[ysimonhan]

What

Adds a new community-maintained Context Hub entry for Langdock Agents Completions. This includes:

• a Python-focused DOC.md
• request format guidance for Langdock's Vercel AI SDK-compatible UIMessage structure
• attachment handling notes
• temporary agent config caveats
• structured output guidance
• a reference note for observed production parsing behavior

Why

Langdock does not currently appear in the public context-hub content tree, and its Agents Completions API has a few integration details that are easy for coding agents to get wrong:

• request messages use Langdock/Vercel-style UIMessage objects rather than OpenAI chat-completions message objects
• uploaded attachments should be referenced via metadata.attachments
• structured output may still return fenced JSON, so production parsers should handle that defensively
• dedicated deployments use a different base URL shape

Testing

• npm test passes
• chub build sample-content/ --validate-only succeeds
• Manual testing done (describe below)

Manual testing:

• validated the new content locally with CHUB_TELEMETRY=0 node cli/bin/chub build content --validate-only
• reviewed the doc structure against existing Context Hub entries
• based implementation details on official Langdock docs plus sanitized observed API behavior from successful Agent Completions usage

Notes

Official sources used:

• https://docs.langdock.com/api-endpoints/api-introduction
• https://docs.langdock.com/api-endpoints/agent/agent
• https://docs.langdock.com/api-endpoints/agent/attachments

The production-notes reference file documents observed behavior from a sanitized real-world batch workflow, but it is phrased as implementation guidance rather than stronger contract guarantees than the official docs