@obtoai/agent-bridge 0.1.0-beta.12 → 0.1.0-beta.13

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@obtoai/agent-bridge",
-  "version": "0.1.0-beta.12",
+  "version": "0.1.0-beta.13",
   "description": "Local consumer for the OBTO Agent Bridge. Receives bridge events over SSE and drives a coding agent (Claude Code or OpenAI Codex) on your machine.",
   "license": "Apache-2.0",
   "author": "OBTO Inc.",

package/src/external-scanner.js

@@ -29,7 +29,8 @@ const CODEX_DIR = path.join(os.homedir(), '.codex', 'sessions');
 const PREVIEW_MAX_CHARS = 200;
 const TITLE_MAX_CHARS = 80;
 const TAIL_READ_BYTES = 8192; // read the last 8KB of each JSONL for the last message
-const HEAD_READ_BYTES = 4096; // read the first 4KB for the first-user-message title
+const TITLE_MAX_LINES = 40;    // scan up to this many lines for a real first message
+const TITLE_MAX_BYTES = 65536; // hard ceiling per file even if MAX_LINES never reached
 
 // Read the tail of a (potentially large) JSONL file without slurping the
 // whole thing into memory. Returns a string (UTF-8) or '' on any failure.
@@ -52,18 +53,76 @@ const readTail = (filePath, maxBytes = TAIL_READ_BYTES) => {
   }
 };
 
-// Read the head of a JSONL file (first ~4KB). Used to extract the first user
-// message for a meaningful session title — "fix the auth bug" beats "OBTO".
-const readHead = (filePath, maxBytes = HEAD_READ_BYTES) => {
+// Claude Code (and to a lesser extent Codex) prepend a parade of injected
+// "user" messages BEFORE the human's first real prompt — system reminders,
+// IDE state, untrusted-metadata wrappers, command blocks, etc. These are
+// noise from a label perspective. Filter them so the first PLAIN user
+// prompt is what we use as the thread title.
+const isInjectionMessage = (text) => {
+  const s = String(text || '').trim();
+  if (!s) return true;
+  if (/^<[a-zA-Z!]/.test(s)) return true;                         // <system-reminder>, <ide_opened_file>, etc.
+  if (/^Sender \(untrusted metadata\)/i.test(s)) return true;
+  if (/^Conversation info \(untrusted metadata\)/i.test(s)) return true;
+  if (/^untrusted metadata/i.test(s)) return true;
+  if (/^Caveat:/i.test(s)) return true;
+  if (/^Claude was launched/i.test(s)) return true;
+  if (/^```json/i.test(s)) return true;
+  if (/^`<command-name>/.test(s)) return true;
+  if (/^This session is being continued/i.test(s)) return true;   // /resume preamble
+  return false;
+};
+
+// Walk a JSONL file line-by-line until we find a plain user message (skipping
+// metadata injections) or we've checked TITLE_MAX_LINES / TITLE_MAX_BYTES.
+// Streaming so big files don't get fully slurped. Returns a clean ≤80-char
+// title or '' if nothing meaningful was found in the bound.
+const extractTitleFromFile = (filePath) => {
   let fd = null;
   try {
-    const stat = fs.statSync(filePath);
-    const len = Math.min(stat.size, maxBytes);
-    if (len <= 0) return '';
     fd = fs.openSync(filePath, 'r');
-    const buf = Buffer.alloc(len);
-    fs.readSync(fd, buf, 0, len, 0);
-    return buf.toString('utf8');
+    const CHUNK = 16 * 1024;
+    const buf = Buffer.alloc(CHUNK);
+    let carry = '';
+    let bytesRead = 0;
+    let linesChecked = 0;
+    while (bytesRead < TITLE_MAX_BYTES && linesChecked < TITLE_MAX_LINES) {
+      const n = fs.readSync(fd, buf, 0, CHUNK, bytesRead);
+      if (n === 0) break;
+      bytesRead += n;
+      carry += buf.toString('utf8', 0, n);
+      const lines = carry.split(/\r?\n/);
+      carry = lines.pop() || '';
+      for (const line of lines) {
+        if (linesChecked >= TITLE_MAX_LINES) break;
+        if (!line.trim()) continue;
+        linesChecked++;
+        let obj;
+        try { obj = JSON.parse(line); } catch (_) { continue; }
+        let role = null, raw = null;
+        if (obj && obj.message && (obj.message.role || obj.type)) {
+          role = obj.message.role || (obj.type === 'user' ? 'user' : 'assistant');
+          raw = obj.message.content;
+        } else if (obj && obj.role && (obj.content || obj.text)) {
+          role = obj.role;
+          raw = obj.content != null ? obj.content : obj.text;
+        }
+        if (role !== 'user' || raw == null) continue;
+        let text = '';
+        if (typeof raw === 'string') text = raw;
+        else if (Array.isArray(raw)) {
+          text = raw
+            .filter((p) => p && (p.type === 'text' || typeof p.text === 'string'))
+            .map((p) => String(p.text || ''))
+            .join(' ');
+        }
+        text = text.replace(/\s+/g, ' ').trim();
+        if (!text) continue;
+        if (isInjectionMessage(text)) continue;
+        return text.length > TITLE_MAX_CHARS ? text.slice(0, TITLE_MAX_CHARS) : text;
+      }
+    }
+    return '';
   } catch (_) {
     return '';
   } finally {
@@ -71,39 +130,6 @@ const readHead = (filePath, maxBytes = HEAD_READ_BYTES) => {
   }
 };
 
-// Walk JSONL lines forward to find the first user message. Same tolerance
-// for the two SDKs' shapes as extractLastMessage. Returns a short title
-// suitable for a thread name, or '' if no user message landed in the head.
-const extractTitle = (jsonlHead) => {
-  if (!jsonlHead) return '';
-  const lines = jsonlHead.split(/\r?\n/).filter((l) => l.trim().length > 0);
-  for (let i = 0; i < lines.length; i++) {
-    let obj;
-    try { obj = JSON.parse(lines[i]); } catch (_) { continue; }
-    let role = null;
-    let raw = null;
-    if (obj && obj.message && (obj.message.role || obj.type)) {
-      role = obj.message.role || (obj.type === 'user' ? 'user' : 'assistant');
-      raw = obj.message.content;
-    } else if (obj && obj.role && (obj.content || obj.text)) {
-      role = obj.role;
-      raw = obj.content != null ? obj.content : obj.text;
-    }
-    if (role !== 'user' || raw == null) continue;
-    let text = '';
-    if (typeof raw === 'string') text = raw;
-    else if (Array.isArray(raw)) {
-      text = raw
-        .filter((p) => p && (p.type === 'text' || typeof p.text === 'string'))
-        .map((p) => String(p.text || ''))
-        .join(' ');
-    }
-    text = text.trim().replace(/\s+/g, ' ');
-    if (text) return text.slice(0, TITLE_MAX_CHARS);
-  }
-  return '';
-};
-
 // Walk a JSONL tail backwards, parse each non-empty line as JSON, return the
 // first one we can extract a message from. Tolerant of multiple shapes —
 // Claude and Codex write slightly different envelopes and the formats have
@@ -196,9 +222,8 @@ const scanClaude = () => {
       let stat;
       try { stat = fs.statSync(filePath); } catch (_) { continue; }
       const tail = readTail(filePath);
-      const head = readHead(filePath);
       const lastMsg = extractLastMessage(tail);
-      const title = extractTitle(head);
+      const title = extractTitleFromFile(filePath);
       out.push({
         source: 'claude',
         sessionId,