fix(email): strip compound channel prefix before SMTP To header

[liuhao1024]

What does this PR do?

Strips the email: platform prefix from compound channel IDs before using them as the SMTP To header. When the cron delivery pipeline or session keys carry a compound channel ID like email:user@example.com, strict SMTP servers (Gmail) reject the message with 555 5.5.2 because email:user@example.com is not a valid RFC 5321 address.

Related Issue

Fixes #56569

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)

Changes Made

• plugins/platforms/email/adapter.py: Added _strip_channel_prefix() helper that strips the email: prefix from channel identifiers. Applied it in all 4 SMTP send paths:
  • EmailAdapter._send_email() — adapter reply path
  • EmailAdapter._send_email_with_attachments() — multi-attachment path
  • EmailAdapter._send_email_with_attachment() — single-attachment path
  • _standalone_send() — standalone/cron delivery path
• tests/gateway/test_email.py: Added TestStripChannelPrefix with 5 tests covering prefix stripping, bare addresses, empty strings, and both SMTP send paths with compound IDs

How to Test

1. Run python -m pytest tests/gateway/test_email.py::TestStripChannelPrefix -xvs — all 5 tests should pass
2. Run python -m pytest tests/gateway/test_email.py -x — full email test suite (93 tests) should pass
3. Verify that msg["To"] receives user@example.com (not email:user@example.com) when a compound channel ID is passed to any SMTP send path

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):)
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature (no unrelated commits)
• I've run pytest tests/gateway/test_email.py -q and all tests pass (93 passed)
• I've added tests for my changes
• I've tested on my platform: macOS

Documentation & Housekeeping

• I've updated relevant documentation — or N/A (internal helper, no user-facing docs)
• I've updated cli-config.yaml.example — or N/A
• I've updated CONTRIBUTING.md or AGENTS.md — or N/A
• I've considered cross-platform impact — N/A (Python stdlib email, no platform-specific code)
• I've updated tool descriptions/schemas — or N/A

Screenshots / Logs

tests/gateway/test_email.py::TestStripChannelPrefix::test_send_email_strips_prefix PASSED
tests/gateway/test_email.py::TestStripChannelPrefix::test_standalone_send_strips_prefix PASSED
tests/gateway/test_email.py::TestStripChannelPrefix::test_strip_email_prefix PASSED
tests/gateway/test_email.py::TestStripChannelPrefix::test_strip_empty PASSED
tests/gateway/test_email.py::TestStripChannelPrefix::test_strip_no_prefix PASSED
93 passed, 1 warning in 3.97s

[alt-glitch]

This was generated by AI during triage.

Related: fixes #56569 (email adapter using the compound email: channel ID as the SMTP To header). Fix PR for that issue — related, not a duplicate.

[teknium1]

Thanks for the localized defensive fix. The stated cron/session-key route is not demonstrated against current main.

Problems

• Direct target parsing removes the outer email: selector before dispatch (tools/send_message_tool.py:361-369, tools/send_message_tool.py:586-589), and explicit cron targets use the same parser (cron/scheduler.py:1128-1159). Inbound email sources are built from the bare extracted sender address (plugins/platforms/email/adapter.py:872-878). The new tests call the low-level helpers directly, so they do not establish the reported production path.
• The diff modifies all three adapter SMTP builders, but tests cover only _send_email and _standalone_send; the attachment helpers have no regression assertion.

Suggested changes

• Add a test that reaches a supported route which actually passes a compound recipient to SMTP, if one exists; otherwise describe this as defensive low-level normalization.
• Cover both attachment send helpers, or centralize the normalization at their common boundary.

Automated hermes-sweeper review.