Current Situation
With the current SwitchBot OpenAPI (v1.1), MindClip recordings only expose recording metadata (title, date/time, duration, status) and the AI-extracted to-do list. The full plain-text transcript — which is visible in the app and can be manually copied/exported (.txt / .docx) from the recording detail screen — is not accessible via the API.
This makes it difficult to build reliable downstream automation, since MindClip's own AI-generated summary occasionally contains factual inaccuracies (e.g. misstating deal terms or dates) that cannot be verified or corrected without access to the source transcript.
Proposed Change
Add an API endpoint (e.g. GET /v1.1/mindclip/recordings/{recordingId}/transcript) that returns the full plain-text transcript of a given recording.
This would let developers build more accurate downstream summarization/analysis on top of MindClip data, instead of relying solely on MindClip's built-in AI summary.
Additional Context
Use case: I'm building an automated daily pipeline that ingests MindClip recordings and re-summarizes them with a different AI model for better accuracy (proper nouns, terminology, etc.). Manually exporting the transcript from the app every day isn't practical for an automated workflow, so API access to the raw transcript would be very valuable — especially for business/professional use cases (we're on the Unlimited plan).
SwitchBot support (ticket #1360922) confirmed the transcript data exists in the app but is not currently exposed via API v1.1, and directed me to file this request here.
Current Situation
With the current SwitchBot OpenAPI (v1.1), MindClip recordings only expose recording metadata (title, date/time, duration, status) and the AI-extracted to-do list. The full plain-text transcript — which is visible in the app and can be manually copied/exported (.txt / .docx) from the recording detail screen — is not accessible via the API.
This makes it difficult to build reliable downstream automation, since MindClip's own AI-generated summary occasionally contains factual inaccuracies (e.g. misstating deal terms or dates) that cannot be verified or corrected without access to the source transcript.
Proposed Change
Add an API endpoint (e.g. GET /v1.1/mindclip/recordings/{recordingId}/transcript) that returns the full plain-text transcript of a given recording.
This would let developers build more accurate downstream summarization/analysis on top of MindClip data, instead of relying solely on MindClip's built-in AI summary.
Additional Context
Use case: I'm building an automated daily pipeline that ingests MindClip recordings and re-summarizes them with a different AI model for better accuracy (proper nouns, terminology, etc.). Manually exporting the transcript from the app every day isn't practical for an automated workflow, so API access to the raw transcript would be very valuable — especially for business/professional use cases (we're on the Unlimited plan).
SwitchBot support (ticket #1360922) confirmed the transcript data exists in the app but is not currently exposed via API v1.1, and directed me to file this request here.