BoltAI2's MCP file tools are thin wrappers around POSIX file APIs. They're designed for simplicity, not scale. The serializer bottleneck (JSON validation before payload size check) is architectural—it affects ANY tool in this MCP server that uses JSON parameter arrays. The serializer limit is a BoltAI2 server constraint, not a file-tool defect. Even if you swapped the underlying tool, you'd hit the same JSON validation ceiling on the MCP protocol layer. Better Strategy Than Tool Swapping: Instead of replacing BoltAI2's tools, adopt a hybrid workflow: For reads >100 kB: Use ripgrep-based search tools (no size limit on results). For writes >1,500 chars: Pre-split into chunks BEFORE calling edit_file (we're already doing this). For bulk file ops: Use the mcp__directory_tree_17ba7w to get the structure first, then target individual small edits. For escape-sensitive paths with spaces: NEVER use bash; always use the file tool wrappers.

Nigel Protter about 6 hours ago

2

🐛

Bug Reports

Description: When using Google Gemini models (gemini-flash-lite-latest, Gemini 2.0 / 2.5) with tool calls enabled (e.g., Web Search via brave_search), streaming responses continuously fail with a Zod invalid_union / invalid_type error. Cause: Google's Gemini OpenAI-compatibility API endpoint drops the index property inside delta.tool_calls[0] when streaming a response. Because BoltAI's streaming parser expects strict OpenAI completion formatting where tool_calls[0].index must be an integer, the schema validation fails and crashes the stream before the function/tool can execute. Raw Error Log: JSON Type validation failed: Value: {"choices":[{"delta":{"role":"assistant","tool_calls":[{"extra_content":{"google":{"thought_signature":"... "}},"function":{"arguments":"{\"query\":\"...\"}","name":"brave_search"},"id":"a4Fsai5e","type":"function"}]},"index":0}],"created":1785262181,"id":"...","model":"models/gemini-flash-lite-latest","object":"chat.completion.chunk","usage":{"completion_tokens":29,"prompt_tokens":651,"total_tokens":680}} Error message: [{"code":"invalid_union","errors":[[{"expected":"number","code":"invalid_type","path":["choices",0,"delta","tool_calls",0,"index"],"message":"Invalid input"}],[{"expected":"object","code":"invalid_type","path":["error"],"message":"Invalid input"}]],"path":[],"message":"Invalid input"}] Suggested Fix: In BoltAI's streaming delta parser schema, make index an optional property inside tool_calls, or fallback to defaulting tool_calls[i].index to 0 when parsing chunks from Google Gemini endpoints.

injectedlinen about 7 hours ago

1

🐛

Bug Reports

Your OS version. Tahoe 26.4.1 Your BoltAI app version. 2.14.0 (build 77) Are you a Setapp user. No. The AI provider & model you’re using. Any. Steps to reproduce the issue. There’s never been a way to save a temporary chat directly from the instant chat bar (ie to be a non-temporary chat), but what did work in previous releases of Bolt 2 was clicking “Open in Companion Window” and then doing Conversation » “Save Temporary Chat.” Now, that doesn’t work either so there is no way to save off a temporary chat once created in the instant chat bar. The error message. NA.

Matt 2 days ago

🐛

Bug Reports

Your OS version. Tahoe 26.4.1 Your BoltAI app version. 2.14.0 (build 77) Are you a Setapp user. No. The AI provider & model you’re using. Any. Steps to reproduce the issue. Control + click on any chat in the left hand side bar. No context menu appears. Renaming is still possible in the right hand side bar and deleting is still possible by shortcut, but there should be a context menu, right? The error message. NA.

Matt 2 days ago

1

🐛

Bug Reports

Your OS version. Tahoe 26.4.1 Your BoltAI app version. 2.14.0 (build 77) Are you a Setapp user. No. The AI provider & model you’re using. OpenRouter. DeepSeek V4 Flash. Steps to reproduce the issue. Start a chat with the instant chat bar. See the video below to see what happens maybe ~20% of the time. The error message. NA.

Matt 2 days ago

🐛

Bug Reports

Your OS version. Tahoe 26.4.1 Your BoltAI app version. 2.14.0 (build 77) Are you a Setapp user. No. The AI provider & model you’re using. Any. Steps to reproduce the issue. I often leave Bolt AI in the background so I can use the instant chat bar. Sometimes I go days without opening the main window. Now, when I do open it, nearly all the time the main menu bar (file, edit, view, etc) is not present. Just blank space. Only killing and reopening the whole app brings it back. The error message. NA.

Matt 2 days ago

3

🐛

Bug Reports

Error undefined is not an object (evaluating 'w.providerMetadata') the models that get loaded automatically like M2.7 do work. related: Minimax also has an oath authentication method, might be useful to add this to boltAI. The same goes for Xai, supergrok subscriptions can not use an API key, only oauth.

atomic004 3 days ago

🐛

Bug Reports

It indicates that it changed, but when I close the lightbox, it’s still the old model. When I open the lightbox again, it shows the new model, but when I close it again, it’s still the old model. When I start a new chat, it’s also still the old model.

philraymond 6 days ago

2

🐛

Bug Reports

Tahoe 26.6 BoltAl 2 Version 2.14.0 (build 77) Not on setup I get random error messages about wanting a master password because my APIkeys are encrypted. 1) I have not encrypted them 2) Sometimes the same model is working and after restaing Bolt it is not ¯|(ツ)/¯

Christian Hofer 7 days ago

3

🐛

Bug Reports

I share the same extract text from image prompt for gemma4 from macOS and iOS, but iOS never “gets” the file.

FossilizedCarlos 9 days ago

💻

Feature Requests

Currently I have added Ollama Cloud to BoltAI 2 as a custom OpenAI provider, howevr it would be nice (cosmetically) if the endpoint https://ollama.com could be added to the Ollama provider, along with an API key so that the Ollama icon is used.

paulgit 11 days ago

💻

Feature Requests

https://cursor.com/en-US/docs/cloud-agent/api/endpoints#create-an-agent describes how to access the Cursor via API-Key. Yet we’re not able to access their models, e.g. composer 2.5. Please add Cursor in the list of AI-Providers

Johannes H 13 days ago

💻

Feature Requests

I installed MacOS Golden Gate and I am now unable to connect to Fireworks or Sakana Fugu. I get an invalid API key error message although the key is absolutely correct. MacOS Golden Gate (27 - Developer Beta) BoltAI 2.13.7 build 76 Not a SetApp User Fireworks or Sakana Fugu (Deep Infra works for some reason)

Matthew Hamilton 13 days ago

2

🐛

Bug Reports

The hanging behavior is a known technical issue in the integration between the local client application (the chat interface or terminal wrapper) and the LLM API. It is called a Tool-Call Loop Hang. Here is the step-by-step sequence that causes the freeze: AI Initiates Tool: The user requests a file modification or read. The AI generates a tool call (e.g., mcp_edit_file_6lpu9l) and stops generating text, as required by the API protocol. MCP Server Executes: The local MCP server successfully executes the file edit and returns the result (the diff or file contents) to the client application. The Integration Gap (The Hang): In a fully automated integration, the client application is supposed to immediately feed the tool's output back to the LLM to trigger the next text generation turn. However, in this specific client setup, the application fails to auto-trigger the next turn. It pauses and waits. The Wake-Up: Because the client is waiting for the AI, and the AI is waiting for the client to send the tool results, the interface hangs. When the user types any message (e.g., "are you stuck again?"), it forces the client application to package the entire conversation history—including the pending tool results—and send it to the LLM. This instantly wakes the AI up, resulting in the immediate "AI-OK" response. 𝐇𝐨𝐰 𝐭𝐨 𝐅𝐢𝐱 𝐨𝐫 𝐌𝐢𝐭𝐢𝐠𝐚𝐭𝐞 𝐭𝐡𝐞 𝐈𝐬𝐬𝐮𝐞 Client-Side Configuration (Permanent Fix): The developer of the client application or terminal wrapper must update the execution logic. The client must automatically post the tool output back to the LLM API endpoint to request the next completion turn without waiting for manual user input. User Workaround (Immediate Fix): Whenever a tool call completes (you will see the green tool response block in the interface) and the interface pauses for more than 3 seconds, type a single character (such as a period . or go) and press Enter. This forces the client to send the tool results to the LLM and resumes execution instantly. AI Mitigation (Combining Steps): The AI will minimize sequential tool calls where possible, combining reads or edits into single multi-file operations to reduce the number of turn transitions.

Nigel Protter 15 days ago

1

🐛

Bug Reports

I have reopen windows when logging in checked during shutdown. But most times, after a restart, the main window doesn’t show when I switch to the app via Cmd+Tab. I believe its the same behavior if I tap on the dock icon. While restarting the app works, it’s starting to get to me since I don’t quit it. Hopefully its a quick fix?

resting 23 days ago

🐛

Bug Reports

I love to review the reasoning behind the output for any given post. I really like the way you have the reasoning/thoughts displayed while the model is processing, but upon completion, I see that the “View Thoughts” section is then constricted to a very small popup window. I’d love to be able to expand the “View Thoughts” section and read it in the main window along with the output. Maintaining the same formatting as it has while streaming would be really cool.

david.sheley 23 days ago

💻

Feature Requests

I’d like a shortcut-triggered palette with type-ahead title search for opening one of my chats. Examples can be found in most IDEs/code editors, e.g. ⌘⇧O in Xcode.

tight.clown 29 days ago

💻

Feature Requests

On initial app open, Bolt opens to an empty/new chat but this chat doesn’t follow the settings (model or tool) saved in Settings. The New Chat settings ARE applied if you tap the trailing new chat icon in the navigation bar and create a new chat. But this makes it sluggish to open Bolt and immediately start typing.

chris 30 days ago

📱

BoltAI Mobile

Lots of more capable agents like coding agents have a compact conversation feature that takes the conversation above it and condenses it down to something that reduces context use. I’ve found that context overflow / poisoning can be a problem in long dialogues. I usually kind of do this manually, have it summarize and then make a new discussion, but a way to compact and then a feature to include the previous N up to and including the most recent compacting would be nice.

Adam Ierymenko about 1 month ago

1

💻

Feature Requests