m365-owa-cli 0.1.0

Agent-first CLI for Microsoft 365 Outlook on the web endpoints

m365-owa-cli

m365-owa-cli is a personal, agent-first Python CLI for Microsoft 365 Outlook workflows through Outlook on the web / OWA internal service endpoints.

The current command surface is calendar-focused, but the package name leaves room for mail and tasks if the same OWA-authenticated endpoint strategy proves useful there. Microsoft Graph is intentionally out of scope.

Examples

m365-owa-cli capabilities
m365-owa-cli schema event
m365-owa-cli auth set-token --connection work
m365-owa-cli auth bookmarklet --connection work --raw
m365-owa-cli auth list-connections
m365-owa-cli events list --connection work --day 2026-04-24
m365-owa-cli events delete --connection work --id AAMk... --confirm-event-id AAMk...

Commands emit JSON by default. Use --pretty where supported for local human-readable output.

Configuration

Tokens are stored as plaintext files under:

~/.config/m365-owa-cli/connections/<connection>.token

Tests and automation can override the config root with M365_OWA_CONFIG_DIR.

Manual Token Capture

Generate a local-only bookmarklet helper:

m365-owa-cli auth bookmarklet --connection work --raw

Create a browser bookmark with the generated value as the URL, open Outlook on the web, click the bookmarklet, then refresh or open Calendar. If OWA sends an Authorization: Bearer ... header to /owa/service.svc, the helper displays it for copying into auth set-token.

Releases

New versions are published to PyPI from GitHub Actions when a v* tag is pushed. The workflow uses PyPI trusted publishing against the pypi environment in this repository, so no long-lived PyPI API token is stored in GitHub.

Before the first release, configure a pending trusted publisher on PyPI:

• PyPI project name: m365-owa-cli
• Owner: okms
• Repository: m365-owa-cli
• Workflow: publish.yml
• Environment: pypi

Release flow:

# bump version in pyproject.toml, commit
git tag vX.Y.Z
git push origin main --tags

The .github/workflows/publish.yml workflow runs tests, builds the source and wheel distributions, checks them with Twine, and publishes them to PyPI. Tag-less runs can also be triggered manually from the Actions tab.