CatchMeUp CLI 0.2.0 — Preview
Pre-releaseCatchMeUp CLI 0.2.0 — Preview
This is an early-access CLI release, not a production-certified Mac installer.
What's new
- One Python command interface for checkout, package, and future Mac executable builds.
- Installable
catchupcommand with nested help, version reporting, strict options,
caller-relative paths,--home,--no-input, and expanded JSON output. - Student/work workflows: course review, meeting preparation, follow-up tracking,
and local PDF/PPTX/TXT/Markdown supporting materials. - Safer configuration handling, overwrite protection, and user data outside installed code.
- Build recipes and automated source, wheel-installation, and Mac-bundle smoke checks.
Install
Install pipx first, then download the wheel and SHA256SUMS from this release.
Verify the wheel's SHA-256 against its entry in SHA256SUMS before installing:
pipx install /absolute/path/to/catchmeup-0.2.0-py3-none-any.whl
catchup --version
catchup setup
catchup todayPython 3.10+ is required for package/source installation. macOS users can explicitly
install missing audio tools with catchup setup --install-audio after installing
Homebrew. Local tasks/materials need neither audio tools nor an API key.
For an existing library, use catchup --home /absolute/path/to/your/library today
or set CATCHMEUP_HOME. No automatic data migration or deletion occurs. Back up
your library before trying a preview release.
Validation and limitations
- 150 local Python tests passed; clean-wheel and Apple Silicon bundle smoke checks
passed with isolated synthetic data, including PDF import. These checks do not
call live AI services, transcribe real recordings, or validate real-device sync. - The attached wheel/source archives are audited to exclude credentials, recordings,
and local course data. Checksums detect mismatched files, not platform notarization. - No PyPI package, Homebrew tap, or signed/notarized Mac installer is published here.
The unsigned/ad-hoc-signed bundle remains a developer build and is not attached. - Microphone capture and WhisperKit are Mac-specific. Windows is not validated.
- Material extraction is text-based, not full visual/OCR understanding. Material
files do not sync to iPhone yet; study and CLI/iPhone feature parity are incomplete. - AI requests may send source text to your chosen provider. Audio syncing is optional
and may upload recordings through the configured shared-folder service.
See the repository README and docs/CLI_DISTRIBUTION.md for setup, architecture,
known limitations, and the remaining stable-release gates.