🤓 Pack Preview/Demo: Claude Async - No More 50s DEADLINE_EXCEEDED Timeout Limits

Showcase
1

Hey Coda Makers! :waving_hand:

With web search and extended thinking, Claude LLM regularly takes 5+ minutes to generate a complete, thorough response. The problem is that Coda (with good intention) closes all API connections after 50s. Maybe you’ve seen it before, the dreaded DEADLINE_EXCEEDED error.

That’s why I built the Claude Async Pack (Demo Doc). It queues prompts for Claude in Supabase, processes them in Vercel, and returns them to Coda via webhook or polling.

This is still just an early preview. Thanks for your feedback!

claude-async-logo 200x200
claude-async-logo 200x200200×200 31.2 KB

Features

  • Extended Thinking
  • Image analysis (up to 100 images URLs)
  • Web Search
  • Citation Formatting
  • Model selection
  • JSON Mode (always returns parsable JSON response)
  • Token and Cost Tracking
  • Webhook Rate limiting (prevents Coda webhook queue overflow)
  • Automatic retries (webhook monitor catches dropped deliveries)
  • Request tracking (full audit trail)
  • Timeout management (undici agent configuration for long requests)

Thanks for reading! Looking forward to your thoughts and feedback! :raising_hands:

3 Likes
2

@Agile_Dynamics you may be particularly intersted in this one (unless you’ve come up with a more streamlined solution)

1 Like
3

an interesting approach, but I wait for a huge Coda AI update that hopefully makes these work arounds unnecessary. cheers, christiaan

2 Likes
4

@Christiaan_Huizer Have they spoken about such an update?

5

hi @Daniel_Robertshaw1 , they talked about an MCP :right_arrow_curving_down:

i hope this will give us some opportunities.

cheers, Christiaan

1 Like