What does this PR do?

On platforms that accept a message send but return no `message_id` (Signal, GitHub webhooks with `github_comment` delivery), `GatewayStreamConsumer` was re-entering the "first send" path on every tool-call boundary (`on_delta(None)`). Each segment reset cleared `_message_id`, `_accumulated`, and `_last_sent_text`, so the next text chunk triggered a fresh `adapter.send()` — posting one platform message per tool call. This is what caused 155 PR comments under a single response.

The fix: when `_message_id == "no_edit"` (the sentinel set by `_send_or_edit` when a platform returns no ID), skip the segment-break reset. State is preserved so all continuation text accumulates and is delivered once via `_send_fallback_final` at stream end.

Platforms with real message IDs (Telegram, Discord, Slack) are unaffected — their `_message_id` is never `"no_edit"` so the reset still fires normally.

Also makes `prompt_toolkit` imports in `hermes_cli/commands.py` optional so gateway and test environments that lack the package can import `resolve_command`, `gateway_help_lines`, and `COMMAND_REGISTRY` without error.

Related Issue

N/A

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 🔒 Security fix
• 📝 Documentation update
• ✅ Tests (adding or improving test coverage)
• ♻️ Refactor (no behavior change)
• 🎯 New skill (bundled or hub)

Changes Made

• `gateway/stream_consumer.py`: guard the segment-break reset with `self._message_id != "no_edit"` — one-line change to the `if got_segment_break:` condition
• `hermes_cli/commands.py`: wrap `prompt_toolkit` imports in a `try/except ImportError` with stub fallbacks so non-CLI environments can import the module
• `tests/gateway/test_stream_consumer.py`: add `test_no_message_id_segment_breaks_do_not_resend` — 3-segment stream on a no-ID platform asserts exactly 2 sends (initial + fallback), not 3

How to Test

1. Configure a webhook delivery target that returns no `message_id` (e.g. `github_comment`)
2. Send a prompt that triggers multiple tool calls in one turn
3. Confirm only one final message is posted, not one per tool boundary

Or run the new test directly:

```
pytest tests/gateway/test_stream_consumer.py::TestSegmentBreakOnToolBoundary::test_no_message_id_segment_breaks_do_not_resend -v
```

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
• 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/ -q` and all tests pass
• I've added tests for my changes (required for bug fixes, strongly encouraged for features)
• I've tested on my platform: Ubuntu 24.04

Documentation & Housekeeping

• I've updated relevant documentation (README, `docs/`, docstrings) — or N/A
• I've updated `cli-config.yaml.example` if I added/changed config keys — or N/A
• I've updated `CONTRIBUTING.md` or `AGENTS.md` if I changed architecture or workflows — or N/A
• I've considered cross-platform impact (Windows, macOS) per the compatibility guide — or N/A
• I've updated tool descriptions/schemas if I changed tool behavior — or N/A

Screenshots / Logs

N/A