bamboo-agent 2026.4.2

A fully self-contained AI agent backend framework with built-in web services, multi-LLM provider support, and comprehensive tool execution
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51

use actix_web::{web, HttpResponse};

use crate::agent::core::Session;
use crate::agent::llm::models::{ContentPart, ImageUrl};
use crate::server::app_state::AppState;

use super::super::ChatImage;

pub(super) async fn append_user_message(
    state: &web::Data<AppState>,
    session: &mut Session,
    message: &str,
    images: Option<&[ChatImage]>,
) -> Result<(), HttpResponse> {
    // Preserve multimodal parts so that preflight hooks (OCR/fallback) and/or multimodal
    // upstream models can use the images.
    if let Some(images) = images.filter(|items| !items.is_empty()) {
        let mut parts = Vec::new();
        // Always include a text part to keep downstream behavior stable.
        parts.push(ContentPart::Text {
            text: message.to_string(),
        });

        for image in images {
            let (_, url) = match state
                .session_store
                .write_image_attachment(session, &image.base64, image.mime_type.as_deref())
                .await
            {
                Ok(result) => result,
                Err(error) => {
                    return Err(HttpResponse::BadRequest().json(serde_json::json!({
                        "error": format!("Failed to store image attachment: {error}")
                    })));
                }
            };
            parts.push(ContentPart::ImageUrl {
                image_url: ImageUrl { url, detail: None },
            });
        }

        session.add_message(crate::agent::core::Message::user_with_parts(
            message.to_string(),
            parts,
        ));
    } else {
        session.add_message(crate::agent::core::Message::user(message.to_string()));
    }

    Ok(())
}