harald/zeroclaw
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 3

feat(channel): add tool status display and configurable message timeout

Browse source 
Show real-time tool execution status in channels with draft support
(e.g. Telegram with stream_mode=partial). During processing, the draft
message shows "Thinking..." and progressively adds tool lines like
"🔧 shell(ls -la)" as tools execute. The final response replaces
all status lines cleanly via finalize_draft.

Also makes the channel message timeout configurable via
agent.channel_message_timeout_secs (default 300s), replacing the
previously hardcoded constant.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
harald 2026-02-21 07:39:39 +01:00
parent 61a998cae3
commit 7df2102d9d
2 changed files with 109 additions and 21 deletions
Download patch file Download diff file Expand all files Collapse all files

9
src/config/schema.rs
View file

@ -253,6 +253,10 @@ pub struct AgentConfig {
pub parallel_tools: bool,
#[serde(default = "default_agent_tool_dispatcher")]
pub tool_dispatcher: String,
/// Timeout in seconds for processing a single channel message (LLM + tools).
/// Default 300s accommodates on-device LLMs (Ollama) which are slower than cloud APIs.
#[serde(default = "default_channel_message_timeout_secs")]
pub channel_message_timeout_secs: u64,
}
fn default_agent_max_tool_iterations() -> usize {
@ -267,6 +271,10 @@ fn default_agent_tool_dispatcher() -> String {
"auto".into()
}
fn default_channel_message_timeout_secs() -> u64 {
300
}
impl Default for AgentConfig {
fn default() -> Self {
Self {
@ -275,6 +283,7 @@ impl Default for AgentConfig {
max_history_messages: default_agent_max_history_messages(),
parallel_tools: false,
tool_dispatcher: default_agent_tool_dispatcher(),
channel_message_timeout_secs: default_channel_message_timeout_secs(),
}
}
}