fix(app-server): truncate tool previews at utf8 boundaries

Author: OnlyYu1996

src/cortex-app-server/src/tools/filesystem.rs

@@ -4,6 +4,7 @@ use std::path::{Path, PathBuf};
 
 use serde_json::{Value, json};
 
+use super::preview::truncate_at_char_boundary;
 use super::types::ToolResult;
 
 /// Read file contents.
@@ -119,7 +120,7 @@ pub async fn write_file(cwd: &Path, args: Value) -> ToolResult {
                 "filename": std::path::Path::new(path).file_name().and_then(|n| n.to_str()).unwrap_or(""),
                 "extension": extension,
                 "size": content.len(),
-                "content_preview": if content.len() > 500 { &content[..500] } else { content }
+                "content_preview": truncate_at_char_boundary(content, 500)
             })),
         },
         Err(e) => ToolResult::error(format!("Failed to write file: {e}")),

src/cortex-app-server/src/tools/mod.rs

@@ -15,6 +15,7 @@ mod definitions;
 mod executor;
 mod filesystem;
 mod planning;
+mod preview;
 mod search;
 mod security;
 mod shell;

src/cortex-app-server/src/tools/preview.rs

@@ -0,0 +1,38 @@
+//! Helpers for building tool output previews.
+
+/// Truncate a string to at most `max_bytes` bytes without splitting a UTF-8 character.
+pub(super) fn truncate_at_char_boundary(input: &str, max_bytes: usize) -> &str {
+    if input.len() <= max_bytes {
+        return input;
+    }
+
+    let mut end = max_bytes;
+    while end > 0 && !input.is_char_boundary(end) {
+        end -= 1;
+    }
+
+    &input[..end]
+}
+
+#[cfg(test)]
+mod tests {
+    use super::truncate_at_char_boundary;
+
+    #[test]
+    fn keeps_ascii_at_requested_byte_limit() {
+        assert_eq!(truncate_at_char_boundary("abcdef", 3), "abc");
+    }
+
+    #[test]
+    fn backs_up_to_utf8_boundary() {
+        let input = format!("{}{}", "A".repeat(499), "\u{4E2D}");
+
+        assert_eq!(input.len(), 502);
+        assert_eq!(truncate_at_char_boundary(&input, 500), "A".repeat(499));
+    }
+
+    #[test]
+    fn returns_full_input_when_under_limit() {
+        assert_eq!(truncate_at_char_boundary("hello", 500), "hello");
+    }
+}

src/cortex-app-server/src/tools/web.rs

@@ -3,6 +3,7 @@
 use serde_json::Value;
 use tokio::process::Command;
 
+use super::preview::truncate_at_char_boundary;
 use super::types::ToolResult;
 
 /// Fetch content from a URL.
@@ -79,9 +80,10 @@ pub async fn fetch_url(args: Value) -> ToolResult {
 
     // Truncate for display if too long
     let truncated = if content.len() > 100_000 {
+        let preview = truncate_at_char_boundary(&content, 100_000);
         format!(
-            "{}...\n[Truncated at 100000 chars, full size: {} chars]",
-            &content[..100_000],
+            "{}...\n[Truncated at 100000 bytes, full size: {} bytes]",
+            preview,
             content.len()
         )
     } else {
@@ -124,7 +126,7 @@ pub async fn web_search(args: Value) -> ToolResult {
             let html = String::from_utf8_lossy(&output.stdout);
             // Simple extraction of text
             let truncated = if html.len() > 10_000 {
-                &html[..10_000]
+                truncate_at_char_boundary(&html, 10_000)
             } else {
                 &html
             };