agent: ditch aiohttp & define REQUESTS_CA_BUNDLE to fix http proxying / trust the self-signed cert from python

examples/agent/Dockerfile.tools

@@ -4,6 +4,7 @@ RUN python -m pip install --upgrade pip && \
     apt clean cache
 
 COPY requirements.txt /root/
+# COPY . /root/
 WORKDIR /root
 RUN pip install -r requirements.txt
 
@@ -13,5 +14,5 @@ COPY ./tools/*.py /root/tools/
 COPY ./squid/ssl_cert/squidCA.crt /usr/local/share/ca-certificates/squidCA.crt
 RUN chmod 644 /usr/local/share/ca-certificates/squidCA.crt && update-ca-certificates
 
-# ENTRYPOINT [ "python" ]
-# CMD ["serve_tools.py"]
+ENTRYPOINT [ "uvicorn" ]
+CMD ["serve_tools:app", "--host", "0.0.0.0", "--port", "8088"]

examples/agent/docker-compose.yml

@@ -32,12 +32,11 @@ services:
     networks:
       - private_net
     environment:
-      - PORT=8088
+      - VERBOSE=1
       - BRAVE_SEARCH_API_KEY=${BRAVE_SEARCH_API_KEY}
+      - REQUESTS_CA_BUNDLE=/usr/local/share/ca-certificates/squidCA.crt
       - http_proxy=http://outgoing_proxy:3128
       - https_proxy=http://outgoing_proxy:3128
-    entrypoint: python
-    command: serve_tools.py
 
     # entrypoint: /usr/bin/bash
     # command: ["-c", "pip install --upgrade gguf && apt update && apt install -y curl && curl https://ochafik.com && pip install gguf"]

examples/agent/serve_tools.py

@@ -12,6 +12,7 @@
 
         uv run examples/agent/serve_tools.py --port 8088
 '''
+import asyncio
 import logging
 import re
 import fastapi
@@ -24,6 +25,11 @@ from tools.fetch import fetch_page
 from tools.search import brave_search
 from tools.python import python, python_tools
 
+# try:
+#     # https://github.com/aio-libs/aiohttp/discussions/6044
+#     setattr(asyncio.sslproto._SSLProtocolTransport, "_start_tls_compatible", True) # type: ignore
+# except Exception as e:
+#     print(f'Failed to patch asyncio: {e}', file=sys.stderr)
 
 verbose = os.environ.get('VERBOSE', '0') == '1'
 include = os.environ.get('INCLUDE_TOOLS')

examples/agent/squid/conf/squid.conf

@@ -23,7 +23,7 @@ http_access allow all
 refresh_pattern -i ($|\.)(files\.pythonhosted\.org|pypi\.org)/.*?\.(whl|zip|tar\.gz)$ 10080 90% 43200 reload-into-ims
 
 # Cache Debian packages
-refresh_pattern \.debian\.org/.*?\.(deb|udeb|tar\.(gz|xz|bz2)$   129600 100% 129600
+refresh_pattern \.debian\.org/.*?\.(deb|udeb|tar\.(gz|xz|bz2))$   129600 100% 129600
 
 # Configure cache 
 cache_dir ufs /var/spool/squid 10000 16 256

examples/agent/tools/fetch.py

@@ -1,6 +1,7 @@
-import aiohttp
+# import aiohttp
 import html2text
 import logging
+import requests
 
 
 async def fetch_page(url: str):
@@ -10,11 +11,16 @@ async def fetch_page(url: str):
 
     try:
         logging.debug(f'[fetch_page] Fetching %s', url)
-        async with aiohttp.ClientSession() as session:
-            async with session.get(url) as res:
-                res.raise_for_status()
-                content = await res.text()
-    except aiohttp.ClientError as e:
+        response = requests.get(url)
+        response.raise_for_status()
+        content = response.text
+        # async with aiohttp.ClientSession(trust_env=True) as session:
+        #     async with session.get(url) as res:
+        #         res.raise_for_status()
+        #         content = await res.text()
+    # except aiohttp.ClientError as e:
+    #     raise Exception(f'Failed to fetch {url}: {e}')
+    except requests.exceptions.RequestException as e:
         raise Exception(f'Failed to fetch {url}: {e}')
 
     # NOTE: Pyppeteer doesn't work great in docker, short of installing a bunch of dependencies

examples/agent/tools/search.py

@@ -1,13 +1,13 @@
-import sys
-from pydantic import Field
-import aiohttp
+# import aiohttp
 import itertools
 import json
 import logging
 import os
-from typing import Annotated, Dict, List
+from typing import Dict, List
 import urllib.parse
 
+import requests
+
 
 def _extract_values(keys, obj):
     values = {}
@@ -66,12 +66,18 @@ async def brave_search(*, query: str) -> List[Dict]:
                 for r in results_of_type:
                     yield _extract_values(keys, r)
 
-    async with aiohttp.ClientSession() as session:
-        async with session.get(url, headers=headers) as res:
+    res = requests.get(url, headers=headers)
     if not res.ok:
-                raise Exception(await res.text())
+        raise Exception(res.text)
+    reponse = res.json()
     res.raise_for_status()
-            response = await res.json()
+    response = res.text
+    # async with aiohttp.ClientSession(trust_env=True) as session:
+    #     async with session.get(url, headers=headers) as res:
+    #         if not res.ok:
+    #             raise Exception(await res.text())
+    #         res.raise_for_status()
+    #         response = await res.json()
 
     results = list(itertools.islice(extract_results(response), max_results))
     print(json.dumps(dict(query=query, response=response, results=results), indent=2))