refactor: use content block helpers and shared url dedupe; run format and lint

apps/agent/agent/integrations/langsmith.py

@@ -9,7 +9,7 @@ import contextlib
 import os
 import time
 from abc import ABC, abstractmethod
-from typing import TYPE_CHECKING, Any
+from typing import Any
 
 from deepagents.backends.protocol import (
     ExecuteResponse,
@@ -19,7 +19,6 @@ from deepagents.backends.protocol import (
     WriteResult,
 )
 from deepagents.backends.sandbox import BaseSandbox
-
 from langsmith.sandbox import Sandbox, SandboxClient, SandboxTemplate
 
 
@@ -148,9 +147,7 @@ class LangSmithBackend(BaseSandbox):
         responses: list[FileDownloadResponse] = []
         for path in paths:
             content = self._sandbox.read(path)
-            responses.append(
-                FileDownloadResponse(path=path, content=content, error=None)
-            )
+            responses.append(FileDownloadResponse(path=path, content=content, error=None))
         return responses
 
     def upload_files(self, files: list[tuple[str, bytes]]) -> list[FileUploadResponse]:
@@ -209,10 +206,7 @@ class LangSmithProvider(SandboxProvider):
                 template_name=resolved_template_name, timeout=timeout
             )
         except Exception as e:
-            msg = (
-                f"Failed to create sandbox from template "
-                f"'{resolved_template_name}': {e}"
-            )
+            msg = f"Failed to create sandbox from template '{resolved_template_name}': {e}"
             raise RuntimeError(msg) from e
 
         # Verify sandbox is ready by polling

apps/agent/agent/middleware/check_message_queue.py

@@ -8,7 +8,6 @@ human messages before the next model call.
 from __future__ import annotations
 
 import logging
-import os
 from typing import Any
 
 import httpx
@@ -38,12 +37,9 @@ async def _build_blocks_from_payload(
 
     if not image_urls:
         return blocks
-    linear_api_key = os.environ.get("LINEAR_API_KEY", "")
     async with httpx.AsyncClient() as client:
         for image_url in image_urls:
-            image_block = await fetch_image_block(
-                image_url, client, linear_api_key=linear_api_key
-            )
+            image_block = await fetch_image_block(image_url, client)
             if image_block:
                 blocks.append(image_block)
     return blocks
@@ -109,17 +105,13 @@ async def check_message_queue_before_model(  # noqa: PLR0911
         content_blocks: list[dict[str, Any]] = []
         for msg in queued_messages:
             content = msg.get("content")
-            if isinstance(content, dict) and (
-                "text" in content or "image_urls" in content
-            ):
+            if isinstance(content, dict) and ("text" in content or "image_urls" in content):
                 logger.debug("Queued message contains text + image URLs")
                 blocks = await _build_blocks_from_payload(content)
                 content_blocks.extend(blocks)
                 continue
             if isinstance(content, list):
-                logger.debug(
-                    "Queued message contains %d content block(s)", len(content)
-                )
+                logger.debug("Queued message contains %d content block(s)", len(content))
                 content_blocks.extend(content)
                 continue
             if isinstance(content, str) and content:

apps/agent/agent/middleware/open_pr.py

@@ -181,16 +181,12 @@ I've {action} pull request to address this issue:
 
         logger.info("Changes detected, preparing PR for thread %s", thread_id)
 
-        current_branch = await asyncio.to_thread(
-            git_current_branch, sandbox_backend, repo_dir
-        )
+        current_branch = await asyncio.to_thread(git_current_branch, sandbox_backend, repo_dir)
 
         target_branch = f"open-swe/{thread_id}"
 
         if current_branch != target_branch:
-            await asyncio.to_thread(
-                git_checkout_branch, sandbox_backend, repo_dir, target_branch
-            )
+            await asyncio.to_thread(git_checkout_branch, sandbox_backend, repo_dir, target_branch)
 
         await asyncio.to_thread(
             git_config_user,

apps/agent/agent/server.py

@@ -4,7 +4,6 @@
 # Suppress deprecation warnings from langchain_core (e.g., Pydantic V1 on Python 3.14+)
 # ruff: noqa: E402
 import logging
-import os
 import warnings
 
 logger = logging.getLogger(__name__)
@@ -29,6 +28,7 @@ from deepagents.backends.protocol import SandboxBackendProtocol
 from langchain_anthropic import ChatAnthropic
 
 from .encryption import decrypt_token
+from .integrations.langsmith import _create_langsmith_sandbox
 from .middleware import (
     ToolErrorMiddleware,
     check_message_queue_before_model,
@@ -37,8 +37,6 @@ from .middleware import (
 )
 from .prompt import construct_system_prompt
 from .tools import commit_and_open_pr, fetch_url, http_request
-from .integrations.langsmith import _create_langsmith_sandbox
-
 
 client = get_client()
 
@@ -46,12 +44,12 @@ SANDBOX_CREATING = "__creating__"
 SANDBOX_CREATION_TIMEOUT = 180
 SANDBOX_POLL_INTERVAL = 1.0
 
-from .utils.sandbox_state import SANDBOX_BACKENDS, get_sandbox_id_from_metadata
 from .utils.github import (
     git_has_uncommitted_changes,
     is_valid_git_repo,
     remove_directory,
 )
+from .utils.sandbox_state import SANDBOX_BACKENDS, get_sandbox_id_from_metadata
 
 
 async def _clone_or_pull_repo_in_sandbox(  # noqa: PLR0915
@@ -87,9 +85,7 @@ async def _clone_or_pull_repo_in_sandbox(  # noqa: PLR0915
     is_git_repo = await loop.run_in_executor(None, is_valid_git_repo, sandbox_backend, repo_dir)
 
     if not is_git_repo:
-        logger.warning(
-            "Repo directory missing or not a valid git repo at %s, removing", repo_dir
-        )
+        logger.warning("Repo directory missing or not a valid git repo at %s, removing", repo_dir)
         try:
             removed = await loop.run_in_executor(None, remove_directory, sandbox_backend, repo_dir)
             if not removed:

apps/agent/agent/tools/commit_and_open_pr.py

@@ -181,9 +181,7 @@ def commit_and_open_pr(
                 "pr_url": None,
             }
 
-        base_branch = asyncio.run(
-            get_github_default_branch(repo_owner, repo_name, github_token)
-        )
+        base_branch = asyncio.run(get_github_default_branch(repo_owner, repo_name, github_token))
         pr_url, _pr_number, pr_existing = asyncio.run(
             create_github_pr(
                 repo_owner=repo_owner,

apps/agent/agent/utils/github.py

@@ -37,24 +37,18 @@ def remove_directory(sandbox_backend: SandboxBackendProtocol, repo_dir: str) ->
     return result.exit_code == 0
 
 
-def git_has_uncommitted_changes(
-    sandbox_backend: SandboxBackendProtocol, repo_dir: str
-) -> bool:
+def git_has_uncommitted_changes(sandbox_backend: SandboxBackendProtocol, repo_dir: str) -> bool:
     """Check whether the repo has uncommitted changes."""
     result = _run_git(sandbox_backend, repo_dir, "git status --porcelain")
     return result.exit_code == 0 and bool(result.output.strip())
 
 
-def git_fetch_origin(
-    sandbox_backend: SandboxBackendProtocol, repo_dir: str
-) -> ExecuteResponse:
+def git_fetch_origin(sandbox_backend: SandboxBackendProtocol, repo_dir: str) -> ExecuteResponse:
     """Fetch latest from origin (best-effort)."""
     return _run_git(sandbox_backend, repo_dir, "git fetch origin 2>/dev/null || true")
 
 
-def git_has_unpushed_commits(
-    sandbox_backend: SandboxBackendProtocol, repo_dir: str
-) -> bool:
+def git_has_unpushed_commits(sandbox_backend: SandboxBackendProtocol, repo_dir: str) -> bool:
     """Check whether there are commits not pushed to upstream."""
     git_log_cmd = (
         "git log --oneline @{upstream}..HEAD 2>/dev/null "
@@ -75,9 +69,7 @@ def git_checkout_branch(
 ) -> bool:
     """Checkout branch, creating it if needed."""
     safe_branch = shlex.quote(branch)
-    checkout_result = _run_git(
-        sandbox_backend, repo_dir, f"git checkout -b {safe_branch}"
-    )
+    checkout_result = _run_git(sandbox_backend, repo_dir, f"git checkout -b {safe_branch}")
     if checkout_result.exit_code == 0:
         return True
     fallback = _run_git(sandbox_backend, repo_dir, f"git checkout {safe_branch}")
@@ -97,9 +89,7 @@ def git_config_user(
     _run_git(sandbox_backend, repo_dir, f"git config user.email {safe_email}")
 
 
-def git_add_all(
-    sandbox_backend: SandboxBackendProtocol, repo_dir: str
-) -> ExecuteResponse:
+def git_add_all(sandbox_backend: SandboxBackendProtocol, repo_dir: str) -> ExecuteResponse:
     """Stage all changes."""
     return _run_git(sandbox_backend, repo_dir, "git add -A")
 
@@ -112,9 +102,7 @@ def git_commit(
     return _run_git(sandbox_backend, repo_dir, f"git commit -m {safe_message}")
 
 
-def git_get_remote_url(
-    sandbox_backend: SandboxBackendProtocol, repo_dir: str
-) -> str | None:
+def git_get_remote_url(sandbox_backend: SandboxBackendProtocol, repo_dir: str) -> str | None:
     """Get the origin remote URL."""
     result = _run_git(sandbox_backend, repo_dir, "git remote get-url origin")
     if result.exit_code != 0:

apps/agent/agent/utils/multimodal.py

@@ -30,26 +30,22 @@ def extract_image_urls(text: str) -> list[str]:
     urls.extend(IMAGE_MARKDOWN_RE.findall(text))
     urls.extend(IMAGE_URL_RE.findall(text))
 
-    deduped = _dedupe_urls(urls)
+    deduped = dedupe_urls(urls)
     if deduped:
         logger.debug("Extracted %d image URL(s)", len(deduped))
     return deduped
 
 
-
 async def fetch_image_block(
     image_url: str,
     client: httpx.AsyncClient,
-    *,
-    linear_api_key: str | None = None,
 ) -> dict[str, Any] | None:
     """Fetch image bytes and build an image content block."""
     try:
         logger.debug("Fetching image from %s", image_url)
         headers = None
         if "uploads.linear.app" in image_url:
-            if linear_api_key is None:
-                linear_api_key = os.environ.get("LINEAR_API_KEY", "")
+            linear_api_key = os.environ.get("LINEAR_API_KEY", "")
             if linear_api_key:
                 headers = {"Authorization": linear_api_key}
             else:
@@ -62,7 +58,13 @@ async def fetch_image_block(
         content_type = response.headers.get("Content-Type", "").split(";")[0].strip()
         if not content_type:
             guessed, _ = mimetypes.guess_type(image_url)
-            content_type = guessed or "application/octet-stream"
+            if not guessed:
+                logger.warning(
+                    "Could not determine content type for %s; skipping image",
+                    image_url,
+                )
+                return None
+            content_type = guessed
 
         encoded = base64.b64encode(response.content).decode("ascii")
         logger.info(
@@ -77,12 +79,6 @@ async def fetch_image_block(
         return None
 
 
-def _dedupe_urls(urls: list[str]) -> list[str]:
-    seen: set[str] = set()
-    deduped: list[str] = []
-    for url in urls:
-        if url in seen:
-            continue
-        seen.add(url)
-        deduped.append(url)
-    return deduped
+def dedupe_urls(urls: list[str]) -> list[str]:
+    deduped: set[str] = set(urls)
+    return list(deduped)

apps/agent/agent/webapp.py

@@ -11,11 +11,12 @@ from typing import Any
 import httpx
 import jwt
 from fastapi import BackgroundTasks, FastAPI, HTTPException, Request
+from langchain_core.messages.content import create_text_block
 from langgraph_sdk import get_client
 
 # Local import for encryption
 from .encryption import encrypt_token
-from .utils.multimodal import extract_image_urls, fetch_image_block
+from .utils.multimodal import dedupe_urls, extract_image_urls, fetch_image_block
 
 logger = logging.getLogger(__name__)
 
@@ -37,7 +38,6 @@ LINEAR_API_KEY = os.environ.get("LINEAR_API_KEY", "")
 X_SERVICE_AUTH_JWT_SECRET = os.environ.get("X_SERVICE_AUTH_JWT_SECRET", "")
 
 
-
 def get_service_jwt_token_for_user(
     user_id: str, tenant_id: str, expiration_seconds: int = 300
 ) -> str:
@@ -78,9 +78,11 @@ LINEAR_TEAM_TO_REPO: dict[str, dict[str, Any] | dict[str, str]] = {
             "open-swe-v3-test": {"owner": "aran-yogesh", "name": "nimedge"},
             "open-swe-dev-test": {"owner": "aran-yogesh", "name": "TalkBack"},
         },
-        "default": {"owner": "aran-yogesh", "name": "TalkBack"}  # Fallback for issues without project
+        "default": {
+            "owner": "aran-yogesh",
+            "name": "TalkBack",
+        },  # Fallback for issues without project
     },
-
     "LangChain OSS": {
         "projects": {
             "deepagents": {"owner": "langchain-ai", "name": "deepagents"},
@@ -93,9 +95,7 @@ LINEAR_TEAM_TO_REPO: dict[str, dict[str, Any] | dict[str, str]] = {
         },
         "default": {"owner": "langchain-ai", "name": "ai-sdr"},
     },
-    "Docs": {
-        "default": {"owner": "langchain-ai", "name": "docs"}
-    },
+    "Docs": {"default": {"owner": "langchain-ai", "name": "docs"}},
 }
 
 
@@ -676,24 +676,15 @@ async def process_linear_issue(  # noqa: PLR0912, PLR0915
         "Please analyze this issue and implement the necessary changes. "
         "When you're done, commit and push your changes."
     )
-    content_blocks: list[dict[str, Any]] = [{"type": "text", "text": prompt}]
+    content_blocks: list[dict[str, Any]] = [create_text_block(prompt)]
     if image_urls:
-        seen_urls: set[str] = set()
-        deduped_urls: list[str] = []
-        for url in image_urls:
-            if url in seen_urls:
-                continue
-            seen_urls.add(url)
-            deduped_urls.append(url)
-        image_urls = deduped_urls
+        image_urls = dedupe_urls(image_urls)
         logger.info("Preparing %d image(s) for multimodal content", len(image_urls))
         logger.debug("Image URLs: %s", image_urls)
 
         async with httpx.AsyncClient() as client:
             for image_url in image_urls:
-                image_block = await fetch_image_block(
-                    image_url, client, linear_api_key=LINEAR_API_KEY
-                )
+                image_block = await fetch_image_block(image_url, client)
                 if image_block:
                     content_blocks.append(image_block)
         logger.info("Built %d content block(s) for prompt", len(content_blocks))

apps/agent/tests/test_multimodal.py

@@ -77,10 +77,7 @@ def test_extract_image_urls_case_insensitive() -> None:
 
 
 def test_extract_image_urls_deduplication() -> None:
-    text = (
-        "Same URL twice: https://example.com/image.png "
-        "and again https://example.com/image.png"
-    )
+    text = "Same URL twice: https://example.com/image.png and again https://example.com/image.png"
 
     assert extract_image_urls(text) == ["https://example.com/image.png"]