fix: propagate real tool_call_id instead of hardcoded placeholder (#73)

The worker (both agent/worker.rs and worker/runtime.rs) was passing the
literal string "tool_call_id" to ChatMessage::tool_result instead of
the actual tool call ID from the LLM response. This breaks
OpenAI-compatible providers that match tool results to their
corresponding calls by ID.

- Add tool_call_id field to ToolSelection struct
- Propagate ToolCall.id through select_tools() into ToolSelection
- Replace all hardcoded "tool_call_id" usages with selection.tool_call_id
- Generate unique IDs for plan-based synthetic selections
- Add test verifying tool_call_id is preserved

Co-authored-by: Yi LIU <[email protected]>
Co-authored-by: Illia Polosukhin <[email protected]>

src/worker/runtime.rs

@@ -313,6 +313,7 @@ Work independently to complete this job. Report when done."#,
                                 parameters: tc.arguments.clone(),
                                 reasoning: String::new(),
                                 alternatives: vec![],
+                                tool_call_id: tc.id.clone(),
                             };
                             self.process_result(reason_ctx, &selection, result);
                         }
@@ -422,7 +423,7 @@ Work independently to complete this job. Report when done."#,
                 );
 
                 reason_ctx.messages.push(ChatMessage::tool_result(
-                    "tool_call_id",
+                    &selection.tool_call_id,
                     &selection.tool_name,
                     wrapped,
                 ));
@@ -436,7 +437,7 @@ Work independently to complete this job. Report when done."#,
             Err(e) => {
                 tracing::warn!("Tool {} failed: {}", selection.tool_name, e);
                 reason_ctx.messages.push(ChatMessage::tool_result(
-                    "tool_call_id",
+                    &selection.tool_call_id,
                     &selection.tool_name,
                     format!("Error: {}", e),
                 ));