Scira AIS
Scira AI

In Progress

0

Actively being built

Nothing here yet

Done

20+

Recently shipped

CompletedBugs

Scira's `form-data` + `bun` Runtime Bug Blocks Sandbox Tools

## Sandbox tools blocked by `form-data` dynamic-require error in `bun` runtime ### Reproduction Every call to any...

CompletedBugs

Grok 4.6 model returns "Error - An error occurred" and is unusable

When selecting Grok 4.6 as the model in Scira, the response immediately fails with the following error...

CompletedBugs

Follow-up: DeepSeek v4 Pro and GLM 5.2 still capping at ~4,096 output tokens

Thanks for moving this to Completed. I wanted to check in on it — I haven't been able to confirm any change in behavior...

CompletedBugs

Shared links on scira.ai are broken — they return

HTTP 500 and the page shows "useMessageScroller must be used within a MessageScroller" in the console. Please fix.

CompletedBugs

Reddit tool issue - no search results

AI models cannot search Reddit even when Reddit is the only data source. See the screenshot below.

CompletedFeatures

Latest Deepseek & Qwen Models?

Are we in the latest versions of the V4 Flash and Pro models that were officially released? The full versions are out...

CompletedBugs

Search Library Isn't Working

When I open “Search Library,” my search history doesn’t appear at all. The following message appears: ``` Couldn't...

CompletedBugs

Unable to Create a New Project

When attempting to create a new project, the following error occurs:§01 DetailsclientTypeErrorMessageBase UI error #45;...

CompletedBugs

Grok 4.5 model returns "Error - An error occurred" and is unusable

When selecting Grok 4.5 as the model in Scira, the response immediately fails with the following error...

CompletedFeatures

Implement message trimming/compaction to prevent context overflow errors in long conversations

When a conversation gets long, some models start returning errors and stop responding. This seems to be because the...

CompletedBugs

Request to Review Maximum Output Context Length for DeepSeek v4 Pro and GLM 5.2

Based on my observations, DeepSeek v4 Pro and GLM 5.2 may be limited to approximately 4,096 output tokens....

CompletedBugs

Can you please solve this issue

This page is appearing pretty often from last two days where model is thinking and calling tools but after some time...

CompletedFeatures

Kimi 3, Model Lag & Dev Support

Is Kimi 3 coming soon? We are getting model updates, but i have noticed some lag. We are still on Opus 4.6 when 4.8 has...

CompletedBugs

Web Search Not Working — AI Returns No Results for Any Query

When I ask Scira to search for something, the AI attempts to use the web search tool but gets no results back. Every...

CompletedBugs

Errors occur when using DeepSeek V4 Pro or GLM 5.2 in multi-turn conversations

As stated in the subject, using DeepSeek V4 Flash with the same context length produces output without any issues, so...

CompletedBugs

Custom instructions not persisted after save and page refresh

When I create or edit “Custom instructions” in Scira, the UI confirms the change (toggle ON, timestamp updated), but...

CompletedFeatures

Thank you for GLM 5.2

Thank you for the recent addition. Is this the thinking model? 

CompletedBugs

Fixing Screen Freezing Issues in Large Context Environments

When interacting with LLMs (particularly Thinking models), after approximately 4–5 rounds of follow-up exchanges,...

CompletedBugs

I can't exit Apps mode.

After switching to Apps mode, I can't exit it. To exit, I need to switch to Multi Agent mode or another mode first, and...

CompletedFeatures

Option to Disable Auto-Scroll to Bottom on Response Completion

I have been using Scira AI and find the service very helpful overall. However, I would like to request one improvement...