v0.6.3: hubspot integration, kb block improvements

[waleedlatif1]

• fix(hubspot): add missing tickets and oauth scopes to OAuth config (fix(hubspot): add missing tickets and oauth scopes to OAuth config #3653)
• fix(knowledge): infer MIME type from file extension in create/upsert tools (fix(knowledge): infer MIME type from file extension in create/upsert tools #3651)
• fix(ui) Live update resources in resource main view (fix(ui) Live update resources in resource main view #3617)
• fix(mothership): mothership-ran workflows show workflow validation errors (fix(mothership): mothership-ran workflows show workflow validation errors #3634)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
docs	Skipped		Mar 18, 2026 6:51pm

[cursor]

You have used all Bugbot PR reviews included in your free trial for your GitHub account on this workspace.

To continue using Bugbot reviews, enable Bugbot for your team in the Cursor dashboard.

[cursor]

You have used all Bugbot PR reviews included in your free trial for your GitHub account on this workspace.

To continue using Bugbot reviews, enable Bugbot for your team in the Cursor dashboard.

[greptile-apps]

Greptile Summary

This PR delivers two fixes: it adds missing HubSpot OAuth scopes (crm.objects.tickets.write, tickets, oauth) needed for ticket write operations, and it improves knowledge base document tools to infer the correct MIME type and filename from the document's file extension rather than always defaulting to text/plain / .txt.

Key changes:

• oauth.ts / utils.ts: Three new HubSpot scopes added with descriptions. The tickets scope is a legacy v1 umbrella scope that overlaps with the newer granular crm.objects.tickets.read/crm.objects.tickets.write scopes already in the list — this may be intentional (dual-API support) but warrants a comment.
• types.ts: New inferDocumentFileInfo helper centralises MIME-type/filename inference behind a whitelist of text-compatible types. Filenames with unrecognised or binary extensions are normalised to .txt / text/plain. A minor edge case with trailing-dot filenames (e.g. report.) could produce a double-dot filename (report..txt).
• create_document.ts / upsert_document.ts: Both tools now call inferDocumentFileInfo and the browser-side base64 path was refactored from a spread-based String.fromCharCode(...bytes) (which would throw a stack overflow near the 1 MB limit) to a safe for loop — a good correctness fix.

Confidence Score: 4/5

• This PR is safe to merge; the changes are targeted fixes with no critical logic errors.
• The knowledge base MIME inference logic is well-structured and the base64 encoding fix removes a real stack-overflow risk. The OAuth change is additive. Minor style concerns exist around the redundant tickets legacy scope and a trailing-dot edge case in filename normalisation, but neither blocks functionality.
• apps/sim/lib/oauth/oauth.ts (redundant legacy scope), apps/sim/tools/knowledge/types.ts (TEXT_COMPATIBLE_MIME_TYPES set and trailing-dot edge case)

Important Files Changed

Filename	Overview
apps/sim/lib/oauth/oauth.ts	Adds three HubSpot OAuth scopes: crm.objects.tickets.write, tickets (legacy v1), and oauth; the tickets scope overlaps with the granular v3 scopes already present.
apps/sim/lib/oauth/utils.ts	Adds human-readable descriptions for the three new HubSpot scopes; descriptions are accurate and consistent with existing entries.
apps/sim/tools/knowledge/types.ts	Introduces inferDocumentFileInfo helper that maps file extensions to MIME types via getMimeTypeFromExtension; minor edge cases with trailing-dot filenames and missing text/yaml variant noted.
apps/sim/tools/knowledge/create_document.ts	Refactors base64 encoding to avoid stack overflow on large buffers and adopts inferDocumentFileInfo for dynamic MIME type/filename resolution; logic is correct.
apps/sim/tools/knowledge/upsert_document.ts	Consistently applies inferDocumentFileInfo in the upsert path, mirroring the create_document changes; no new issues introduced.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A["inferDocumentFileInfo(documentName)"] --> B["getFileExtension(documentName)"]
    B --> C{ext is non-empty?}
    C -- No --> F["base = documentName"]
    C -- Yes --> D["getMimeTypeFromExtension(ext)"]
    D --> E{mimeType in TEXT_COMPATIBLE_MIME_TYPES?}
    E -- Yes --> G["Return { filename: documentName, mimeType }"]
    E -- No --> H["base = documentName.slice(0, lastDot)"]
    H --> I{base is empty?}
    I -- Yes --> J["use documentName as base"]
    I -- No --> K["use base"]
    F --> L["Return { filename: base + '.txt', mimeType: 'text/plain' }"]
    J --> L
    K --> L

Loading

_{Last reviewed commit: "fix(mothership): mot..."}