`subscribe_outbound()`, `dispatch_outbound()`, and `stop()` have zero
callers — `ChannelManager._dispatch_outbound()` handles all outbound
routing via `consume_outbound()` directly. Remove the dead methods and
their unused imports (`Callable`, `Awaitable`, `logger`).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Inject cache_control: {"type": "ephemeral"} on the system message and
last tool definition for providers that support prompt caching. Adds
supports_prompt_caching flag to ProviderSpec (enabled for Anthropic only)
and skips caching when routing through a gateway.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When a shell command times out, process.kill() is called but the
process object was never awaited after that. This leaves subprocess
pipes undrained and file descriptors open. If many commands time out,
fd leaks accumulate.
Add a bounded wait (5s) after kill to let the process fully terminate
and release its resources.
Address review feedback:
- Remove on_progress call for interim text to prevent duplicate
messages when the model simply answers a direct question
- Reset final_content to None before continue to avoid stale
interim text leaking as the final response on empty retry
Closes#705
- Add image upload via im.v1.image.create API
- Add file upload via im.v1.file.create API
- Support sending images (.png, .jpg, .gif, etc.) as image messages
- Support sending audio (.opus) as voice messages
- Support sending other files as file messages
- Refactor send() to handle media attachments before text content
The send() payload contains user message content (msg.content) which
may include non-ASCII characters (e.g. CJK, German umlauts, emoji).
The auth frame and Discord heartbeat/identify payloads are left
unchanged as they only carry ASCII protocol fields.
Some LLM providers (MiniMax, Gemini Flash, GPT-4.1, etc.) send an
initial text-only response like "Let me investigate..." before actually
making tool calls. The agent loop previously broke immediately on any
text response without tool calls, preventing these models from ever
using tools.
Now, when the model responds with text but hasn't used any tools yet,
the loop forwards the text as progress to the user and gives the model
one additional iteration to make tool calls. This is limited to a
single retry to prevent infinite loops.
Closes#705
