Backlog
18Under consideration
MCP error: add_url_source via MCP never triggers crawler (sources stuck PROCESSING)
Date: 2026-04-05 MCP Server: Chipp (build.chipp.ai/mcp, streamable-http transport) MCP Client: Claude Code v2.1.92 (Anthropic CLI, macOS) Steps to reproduce Call add_url_source via MCP to add a URL-based knowledge source: { "appId": 10041051, "url": "https://portal-conneva.vercel.app/api/mosaic/knowledge/index.html", "crawlAllLinks": true } Result Source is created successfully and returns a source ID (e.g., dd18f40d-5de9-46a1-a1c8-2a41a9686237). However, the source remains in PROCESSING status indefinitely — the crawler never picks it up. Calling get_crawl_progress on the source returns: { "error": { "code": "INTERNAL_ERROR", "message": "Failed to connect before the deadline" } } Expected Source transitions through PENDING → PROCESSING → COMPLETED as the crawler fetches and indexes the content. This is the behavior observed when adding the same URL via the Chipp dashboard. Root cause evidence Dashboard-added sources crawl fine. The same URL was added manually through the build.chipp.ai dashboard and crawled to COMPLETED status successfully. Two individual topic URLs were also added via dashboard and indexed correctly. MCP-created sources never start crawling. Multiple sources created via add_url_source MCP tool remain stuck at PROCESSING. No crawl activity is observed. The crawler itself is healthy. Dashboard-initiated crawls complete normally, confirming the crawler service is running. The issue is that MCP-created sources are not being enqueued for crawling. get_crawl_progress also fails for MCP-created sources with a "Failed to connect before the deadline" error, suggesting the crawl was never initiated (no Temporal workflow to query). Tested multiple times across different sessions and URLs — consistently reproducible. Workaround Add sources manually via the build.chipp.ai dashboard instead of the MCP add_url_source tool. Environment MCP Server: Chipp (build.chipp.ai/mcp, streamable-http, OAuth 2.0 + PKCE) MCP Client: Claude Code v2.1.92 (Anthropic CLI, macOS darwin) All other MCP tools (app management, list_knowledge_sources, delete_knowledge_source, custom actions, system prompt, branding) function normally for this workspace and app.
Email/Codes to sign in aren't working
I am trying to create an account. I can’t login via email or with it sending me a code. When I try to login via email, it directs me to use the code login access. When I request login access via the code it gives me this response: “ That code isn't valid or has expired. Request a new one.” Help with this would be appreciated.
Can`t sugn up
I tried to sign up on your website using both OAuth and the login/password approach. I received an email with a 6-digit code. However, when I try to sign in, I am redirected to the page https://build.chipp.ai/#/login?error=payment_required
Microsoft sign in and Entra for HQs and Apps
We're aware Chipp supports Bring‑Your‑Own‑Auth (a custom-hosted login page that transfers a token to Chipp), but that requires us to build and host our own auth layer. What we're asking for is native Microsoft/Entra integration, which we'd break into three levels — ordered from most to least needed: "Sign in with Microsoft" (most needed): A generic Microsoft sign-in option available to users of the HQ and Apps, similar to the existing Google sign-in option — no connection to our specific Entra tenant required, just a standard Microsoft account login option alongside existing sign-in methods. Direct Entra tenant connection for group-based access control and provisioning: Ability to connect our specific Entra tenant to Chipp HQ and individual apps for more granular control — specifically, mapping Entra/AD groups to determine which apps a user can see and access, and potentially using Entra for user provisioning (e.g., SCIM or a similar sync mechanism) so access can be managed centrally from our directory rather than manually inside Chipp. Note: we use an RBAC structure in AD where users are members of role-based security groups, which are then nested inside resource-based security groups (rather than being added to resource groups directly). If group mapping is supported, it would need to account for nested/transitive group membership, not just direct membership, in order to work with our setup. Multi-tenant support (lowest priority): Ability to connect more than one Entra tenant to the same Chipp workspace, since our organization spans multiple tenants. This is something we could work around in the near term if needed. We see these as three levels of the same overall request, and would value each one, but wanted to be clear about relative priority in case they're not all feasible at once.
Conversation Message Limit
We have a high message per user/conversation. As this is a free service (chatbot) we want to have the ability to ‘limit’ conversations that go over a determined amount. To have a configurable message that says something like “You have reached the message limit for the day, come back tomorrow”. This limit can be a cost or message counter, so don’t go above $10 per conversation or more than 30 messages. I like the cost option. The reason is that we recently had a user have a $50 conversation in one session.
Chip says I’m out of usage credits
Chipp says I’m out of usage and thinks I’m on the free plan but I’m actually on the $30 per month plan. I’m using chip on a browser on my phone.
"Mic access denied" message covers text in input box
After denying microphone access, a "Mic access denied" tooltip appears and overlaps the text input field, making it hard to see/type the message underneath: Could you please reposition the message so it doesn't cover the input box?
Highlight/select color too bright in light mode
The chatbot's default message highlight/select color is very bright/high-contrast in light mode, making it hard to read comfortably: This is especially important for users who want to copy/paste answers, which we do to help support users. Dark mode looks better by comparison, but still seems a bit bright. Could the highlight color be adjusted or muted for light mode, or made theme-aware?
Notification Badge is Clipped by Parent Container
The unread notification badge (blue dot) on the chatbot launcher is being clipped by the parent container's bounding box, which results in a truncated/cut-off appearance. Could you please fix this so the notification dot should renders completely intact, without any clipping? Something like this: Thank you!
Chat Starting Message - More Functionality
Current Starting message only allows for plain text. Is there any way to have more options to allow for hyperlinks. This is what we would like: By continuing, you agree to our terms of use and privacy policy. Instead of: By continuing, you agree to our terms of use (https://www.biblica.com/terms-of-use/) and privacy policy (https://www.biblica.com/privacy-policy/).
Next up
28Committed and queued
API features to manage meta data
I would be great if an API could be used to set users to “subcribed” or revoke this status, to generally query the user database or to trigger workflows when a new user singed up.
Microsoft Teams Deployment
Same as Chipp App Slack deployment but for MS Teams
API access to POST and DELETE knowledge base files
If knowledge base files could be DELETEd and POSTed via API, then this would avoid the manual process of navigating to the Chipp website and clicking to delete and upload, doing this for each bot. That would make a massive difference to my use case, given my large number of resource files multiplied by a large number of Chipp apps.
Unique subdomain for HQ
Would it be possible to add the custom subdomain for your HQ like you do now for the individual Chipps?
Provide token usage details/context
I am troubleshooting a Slack integration, and just received the following error: “Token usage limit exceeded. You've used 270,464 of 100,000 tokens this month.” I’ve reviewed my Settings, the app logs, and other areas in Chipp to determine what is causing my requests to have this higher token usage, but I don’t have a clear answer yet. Can the token usage be added to the Chat Logs as an additional, optional column? Or perhaps have a token usage log, if there are other ways tokens might be used per account? Perhaps I would benefit from creating separate apps with more specific purposes to get the usage per request lower. I can’t be sure without more contextual data, though.
Refresh of knowledge source
One of our knowledge sources is a google sheet from the chats of another faithbot. By default, it is set to “Don’t Refetch.” I can change it to refetch daily or weekly. When I close it out and open it back up again, it is not saved and says - Don’t refetch. How can I make it where it actually saves this setting? As it is, my data hasn’t been refetched in 31 days.
Collect and show multilingual Title, description, and conversation starters.
TL;DR: Please read the browser’s language setting and help us to provide (and present )localized title, description, and conversation starters when they open the chat. User Story: A French speaker accesses our website, using the French interface for the site. They find the chat button and open it. They are presented with a chipp bot’s title, description, and starter questions, all in English. The chat agent is capable of speaking with them in French, but unless they “try” asking a question in French, they may leave without getting the response. Background: We work across thousands of languages, but our sites and software support 10ish "gateway" languages like French, Arabic, Spanish that people usually have to learn to use their phones or computers. Modern localized websites can read the language or location of the browser to guess what language the person might want to use. Most of our software documentation is in English. The goal for our bots is to harness those to provide "bring your own language" support for software through any the 90+ languages that the models can use. Basically, we want it to support the user in “any” language of wider communication, but we have no way to visually communicate to the user that their language is supported. The agents can seamlessly give responses across those languages, except the user has to know to ask the first question in their language. Please give us a way to let the user know that their language is supported. User Side Try to read the browser language setting of the user and present the starting interface (title, description, conversation starter questions, starter message) in that language. The content can be generated or developer-provided. For Right-to-Left languages like Arabic and Hebrew, adjust text positioning accordingly. Even if the site reads the browser settings, It is still helpful to give a a dropbox where the user can override with a “starting/interface” language, and provide the initial interface in that language. Developer Side I would love to translate (and use) 3 strings and 4 questions for each language for each agent to present something more accessible on the front page. Once the user asks the first question, the bot can take over and translate content on the fly. We don’t need to translate the bot’s sources. Options Give the developer the option to define Title, Description and Conversation Starters in a selection of languages. This is the way the Google Play store works. For any languages where the dev hasn’t translated the text, generate localzed versions of the content. Allow the Dev to pass a language flag to the app based on user/browser settings. (Example: the user has selected already the French interface on our website, so we pass “fr” when loading the widget.) This could be passed as a variable when loading the iframe (either as a variable “lang=fr”, or url “https://keymansupport-10011648.chipp.ai/fr”). Alternate Option Suggested I only need to translate (and use) title, description and conversation starters to present something more accessible on the agent’s front page. Valentina has suggested duplicating the bots and providing separate copies for each target language. This presents 2 problems: Even 10 monolingual copies of our 50 bots would push us to 500 bots and become nearly impossible to manage updates to every copy. With a global audience, we don’t know what language the user will want, though we can pick a few obvious targets, and we don’t want to spend time or resources creating a fully separate bot for a language that may never be used.
API access to editing System Prompts
It would be great to be able to GET, DELETE and POST (update) the Prompt for a Chipp chat. If I edit (at my end) master snippets of prompt instructions that are common to many Chipp apps, but then other parts of the prompt vary from app to app, then I can calculate/compile each prompt from its components and send it to Chipp. This would allow the updating and managing of many prompts at once, so that all chats benefit from prompt improvements.
many-to-many mapping of users to apps!
I need a many-to-many mapping of users to apps. In the HQ environment you can DUPLICATE apps from one WORKSPACE to another, but I want to instead LINK the same app to multiple HQs, editing it just in MY workspace. (I need apps A, B, C and F accessible to Group 1, and apps B, C and D accessible to Group 2. Can this be done with HQs? Duplication isn't the right approach, because it forks the version, making hand-edits necessary for every clone.)
Change email in admin workspace.
Hello, Not sure this is a feature request or bug. We cannot change our admin login email in the setting backed. Please advise.
In Progress
8Actively being built
Integrate Chipp with Airtable
I have a lot of data in Airtable, I would like to ask questions about it on Chipp
Chipp as MCP Server
Implement Model Context Prototcol to enable models and assistants to securely access external data sources, tools, and prompts through a standardized client-server architecture
Transcript of voice mode conversations
It would be great if there was a setting to enable automatic transcripts of voice mode conversations. Otherwise, everything is gone after closing the conversation and the feature has limited value.
More Usage & Credit Consumption Analytics
Now with the usage-based pricing, it would be great to have even deeper insight into the credit consumption. E.g. a breakdown by app or even by end-user, number of messages per period, or a usage timeline to identify peaks and spikes. In the long-run it might be cool to also have end-user facing analytics about their own usage, so we could come up with usage-based pricing ourselves.
Discord and Chatbots
Can you create an easy way to link a chatbot to a Discord channel? In the meantime, can you post the instructions for the not so easy way to link them ;)
Source do not load
Sources used do not load when citing - The name is never loaded, and just an ID appears for the citation, it even looks like it’s perpetually still loading based on the animation on the citation. When you click on the citation, you get a blank modal.
Per User usage data not shown
Hi there, in my usage analytics, the per user breakdown doesn’t show data. The users have to sign up for the app.
Youtube imports time out
This seems to happen regularly. I’ve mentioned it in discord and it gets fixed, only to break again. All Youtube imports time out. They seem to do the extraction step but time out on transcription.
Done
61Recently shipped
Please add a way to toggle a widget to nearly fullscreen
Suggestion: Add an expand/contract button (for media over 700px) or pop out/pop in button (like Gmail does with emails) that can toggle a widget to nearly fullscreen. Alternately, give a button to “pop out” the chat to a new fullscreen item. It could look something like this on desktop (hacked via CSS): The iFrame embed code seems to give the option of defining an internal width, but I don’t see such a custom option for widgets. We could override the default width via CSS, but that is fragile and not dynamic.
No URL retrieving from my documents
I don’t know why, but this keeps stopping from working. I uploaded my PDF with links as a source, but when I ask in the chat for a meditation, the app gives me a non-existent link. It worked for a while, but now it doesn’t anymore. I also tried it with an Excel sheet, a Google Sheet, and that doesn’t work either. I don’t understand it at all, and I’ve been trying to get this working for a year already. Is this just never going to work? This is my app: https://mediteren.frankbruining.com
Missing and wrong Chatlogs
I am trying to look up the chatlogs of a specific user but when I enter the user’s email address to the search bar in the “Chats” tab, it doesn’t return anything, even though credits have been consumed and I see records in the synced Google Sheets. When I copy paste some text from the Google Sheets log, I can find the conversation in Chipp. However, in the Google Sheet are different messages compared to what I see in Chipp. In Chipp it seems to be right but in the Google Sheet, I can see messages that apparently were never part of this conversation.
Image recognition accuracy
I think I read somewhere that Chipp had turned on OpenAI's image recognition as default. However, is that right? When configuring an app, the switch to turn it on is still described as using Gemini. I just uploaded a Maths textbook image of several graphs from my phone to a Chipp app and also natively to Chat-GPT. Chat-GPT read it correctly, but the Chipp equivalent got some of it wrong. What’s the best recognition we can get and how?
Docs stuck in processing
My training documents aren’t uploading. Stuck in processing for hours. Any resources?
Cite sources and Cross-Chipp chat conflict in passed URLs
We have a “master” consultant bot that is instructed to ask questions of a bunch of specific software support bots and return those answers. If I turn on “cite sources” on the master bot, Chipp seems to be appending some internal id number to some or all of the hyperlinks passed from the other bots. https://software.sil.org/flextrans/ is correct, but not https://software.sil.org/flextrans/[8785939] Disabling Cite sources on the master bot fixes the error. For more details and to get started, visit the official FLExTrans page: https://software.sil.org/flextrans/[8785939]. This answer is provided by the SIL tools support system.
Chat History offered but not available.
Under the ☰ button, I see the enticing option Chat History on Widget and Standalone views. https://software.sil.org/fieldworks/ I can never get it to save/show history in widget mode by chatting, triggering a new chat, and opening the history. Sometimes this process works in widget mode, but mostly it doesn’t.
User remained subscribed after cancelling
When a paid user cancels the subscription through Stripe. It seems that this does not sync back to Chipp, so the users were still able to use the app without paying and I had to manually set them to “not subscribed” in Chipp.
Can't delete apps
I cannot delete any app in my workspace. It throws an error. Steps to reproduce: Select an app Click the three dots Click delete Type in “delete” Hit delete It throws this error:
Error message "LiveKit configuration missing"
Configuration: Voice enabled. Was working for a day (1-6-26) but today when I click “Enable microphone” I am getting this error message and voice isn’t working.