fix(chat/tools): remove hard output truncation from sandbox and web_fetch

The SDK already handles oversized tool output by writing to tool-results
files and reading back via MCP. Our 50K char truncation was cutting off
output before the agent could see it — the SDK's mechanism is the proper
way to handle large results.

autogpt_platform/backend/backend/api/features/chat/tools/sandbox.py

@@ -16,8 +16,6 @@ import shutil
 
 logger = logging.getLogger(__name__)
 
-# Output limits — prevent blowing up LLM context
-_MAX_OUTPUT_CHARS = 50_000
 _DEFAULT_TIMEOUT = 30
 _MAX_TIMEOUT = 120
 
@@ -253,8 +251,8 @@ async def run_sandboxed(
             stdout_bytes, stderr_bytes = await asyncio.wait_for(
                 proc.communicate(), timeout=timeout
             )
-            stdout = stdout_bytes.decode("utf-8", errors="replace")[:_MAX_OUTPUT_CHARS]
-            stderr = stderr_bytes.decode("utf-8", errors="replace")[:_MAX_OUTPUT_CHARS]
+            stdout = stdout_bytes.decode("utf-8", errors="replace")
+            stderr = stderr_bytes.decode("utf-8", errors="replace")
             return stdout, stderr, proc.returncode or 0, False
         except asyncio.TimeoutError:
             proc.kill()

autogpt_platform/backend/backend/api/features/chat/tools/web_fetch.py

@@ -19,7 +19,6 @@ logger = logging.getLogger(__name__)
 
 # Limits
 _MAX_CONTENT_BYTES = 102_400  # 100 KB download cap
-_MAX_OUTPUT_CHARS = 50_000  # 50K char truncation for LLM context
 _REQUEST_TIMEOUT = aiohttp.ClientTimeout(total=15)
 
 # Content types we'll read as text
@@ -141,16 +140,12 @@ class WebFetchTool(BaseTool):
         if extract_text and "html" in content_type.lower():
             text = _html_to_text(text)
 
-        truncated = len(text) > _MAX_OUTPUT_CHARS
-        if truncated:
-            text = text[:_MAX_OUTPUT_CHARS]
-
         return WebFetchResponse(
-            message=f"Fetched {url}" + (" (truncated)" if truncated else ""),
+            message=f"Fetched {url}",
             url=response.url,
             status_code=response.status,
             content_type=content_type.split(";")[0].strip(),
             content=text,
-            truncated=truncated,
+            truncated=False,
             session_id=session_id,
         )