mirror of
https://github.com/usestrix/strix.git
synced 2026-08-19 10:05:12 +02:00
revert(proxy): drop overfit Caido reconnect/HTTPQL band-aids, keep serialization lock (#799)
Co-authored-by: Ahmed Allam <ahmed39652003@gmail.com>
This commit is contained in:
co-authored by
Ahmed Allam
parent
e4548cb28c
commit
96ca7e544d
@@ -434,6 +434,7 @@ SPECIALIZED TOOLS:
|
||||
PROXY & INTERCEPTION:
|
||||
- Caido CLI - Modern web proxy (already running). Use the proxy tools
|
||||
directly, or import `caido_api` from sandbox Python scripts.
|
||||
- HTTPQL filters (for `list_requests`): quote string values, leave integers unquoted (`resp.code.eq:200`, not `"200"`); combine terms with `AND`/`OR` (there is no `NOT` — use the negated operator `ne`/`ncont`/`nregex`). Numeric fields (`resp.code`, `req.port`) use `eq`/`ne`/`gt`/`gte`/`lt`/`lte`; text fields (`req.host`, `req.path`, `req.method`, `req.raw`) use `cont`/`ncont`/`eq`/`regex`. Example: `resp.code.gte:200 AND resp.code.lt:300 AND req.host.cont:"api"`.
|
||||
- NOTE: If you are seeing proxy errors when sending requests, it usually means you are not sending requests to a correct url/host/port.
|
||||
- Ignore Caido proxy-generated 50x HTML error pages; these are proxy issues (might happen when requesting a wrong host or SSL/TLS issues, etc).
|
||||
|
||||
|
||||
@@ -282,10 +282,6 @@ async def run_strix_scan(
|
||||
context: dict[str, Any] = {
|
||||
"coordinator": coordinator,
|
||||
"sandbox_session": bundle["session"],
|
||||
# One ``SharedCaidoClient`` is reused by every agent in the scan
|
||||
# (child contexts are shallow copies via ``dict(parent_ctx)``). It
|
||||
# serializes access to the non-concurrency-safe GraphQL transport
|
||||
# and rebuilds it if it dies mid-scan.
|
||||
"caido_client": bundle["caido_client"],
|
||||
"agent_id": root_id,
|
||||
"parent_id": None,
|
||||
|
||||
@@ -80,72 +80,30 @@ async def _login_as_guest(
|
||||
raise RuntimeError(f"loginAsGuest failed after {attempts} attempts: {last_err}")
|
||||
|
||||
|
||||
async def _aclose_quietly(client: Client) -> None:
|
||||
"""Best-effort close of a client whose setup failed; never raises."""
|
||||
with contextlib.suppress(Exception):
|
||||
await client.aclose()
|
||||
|
||||
|
||||
async def _connect_client(
|
||||
session: BaseSandboxSession,
|
||||
*,
|
||||
host_url: str,
|
||||
container_url: str,
|
||||
) -> Client:
|
||||
access_token = await _login_as_guest(session, container_url=container_url)
|
||||
client = Client(host_url, auth=TokenAuthOptions(token=access_token))
|
||||
await client.connect()
|
||||
return client
|
||||
|
||||
|
||||
async def bootstrap_caido(
|
||||
session: BaseSandboxSession,
|
||||
*,
|
||||
host_url: str,
|
||||
container_url: str,
|
||||
) -> tuple[Client, str]:
|
||||
"""Connect to the in-container Caido sidecar and select a fresh project.
|
||||
|
||||
Returns the connected client and the id of the temporary project it
|
||||
selected. The project id lets :func:`reconnect_caido` rebuild a dead
|
||||
transport while staying on the *same* project (and its captured traffic)
|
||||
instead of creating a new empty one.
|
||||
"""
|
||||
) -> Client:
|
||||
"""Connect to the in-container Caido sidecar and select a fresh project."""
|
||||
logger.info("Bootstrapping Caido client (host=%s, container=%s)", host_url, container_url)
|
||||
|
||||
client = await _connect_client(session, host_url=host_url, container_url=container_url)
|
||||
access_token = await _login_as_guest(session, container_url=container_url)
|
||||
|
||||
client = Client(host_url, auth=TokenAuthOptions(token=access_token))
|
||||
await client.connect()
|
||||
|
||||
try:
|
||||
project = await client.project.create(
|
||||
CreateProjectOptions(name="sandbox", temporary=True),
|
||||
)
|
||||
await client.project.select(project.id)
|
||||
except BaseException:
|
||||
# Don't leak the connected transport if project setup fails.
|
||||
await _aclose_quietly(client)
|
||||
# The connected client never reaches the session bundle if project
|
||||
# setup fails, so close it here to avoid leaking the transport.
|
||||
with contextlib.suppress(Exception):
|
||||
await client.aclose()
|
||||
raise
|
||||
logger.info("Caido project selected: %s", project.id)
|
||||
return client, str(project.id)
|
||||
|
||||
|
||||
async def reconnect_caido(
|
||||
session: BaseSandboxSession,
|
||||
*,
|
||||
host_url: str,
|
||||
container_url: str,
|
||||
project_id: str,
|
||||
) -> Client:
|
||||
"""Rebuild a Caido client after its transport died, keeping the project.
|
||||
|
||||
Re-authenticates, reconnects, and re-selects the existing project so the
|
||||
caller keeps access to the traffic captured before the disconnect.
|
||||
"""
|
||||
logger.info("Reconnecting Caido client (host=%s, project=%s)", host_url, project_id)
|
||||
client = await _connect_client(session, host_url=host_url, container_url=container_url)
|
||||
try:
|
||||
await client.project.select(project_id)
|
||||
except BaseException:
|
||||
# A missing/unavailable project must not leave the freshly-connected
|
||||
# transport dangling — otherwise every retry leaks another one.
|
||||
await _aclose_quietly(client)
|
||||
raise
|
||||
return client
|
||||
|
||||
@@ -5,20 +5,15 @@ from __future__ import annotations
|
||||
import logging
|
||||
import shutil
|
||||
from pathlib import Path
|
||||
from typing import TYPE_CHECKING, Any
|
||||
from typing import Any
|
||||
|
||||
from agents.sandbox.entries import BaseEntry, LocalDir
|
||||
from agents.sandbox.manifest import Environment, Manifest
|
||||
|
||||
from strix.config import load_settings
|
||||
from strix.runtime.backends import get_backend
|
||||
from strix.runtime.caido_bootstrap import bootstrap_caido, reconnect_caido
|
||||
from strix.runtime.caido_bootstrap import bootstrap_caido
|
||||
from strix.runtime.local_dir_staging import stage_symlink_safe_dir
|
||||
from strix.tools.proxy.caido_api import SharedCaidoClient
|
||||
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from caido_sdk_client import Client as CaidoClient
|
||||
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
@@ -136,24 +131,16 @@ async def create_or_reuse(
|
||||
host_caido_url = f"{scheme}://{caido_endpoint.host}:{caido_endpoint.port}"
|
||||
logger.debug("Caido host endpoint resolved: %s", host_caido_url)
|
||||
|
||||
caido_client, caido_project_id = await bootstrap_caido(
|
||||
caido_client = await bootstrap_caido(
|
||||
session,
|
||||
host_url=host_caido_url,
|
||||
container_url=container_caido_url,
|
||||
)
|
||||
|
||||
async def _reconnect_caido() -> CaidoClient:
|
||||
return await reconnect_caido(
|
||||
session,
|
||||
host_url=host_caido_url,
|
||||
container_url=container_caido_url,
|
||||
project_id=caido_project_id,
|
||||
)
|
||||
|
||||
bundle = {
|
||||
"client": client,
|
||||
"session": session,
|
||||
"caido_client": SharedCaidoClient(caido_client, _reconnect_caido),
|
||||
"caido_client": caido_client,
|
||||
}
|
||||
_SESSION_CACHE[scan_id] = bundle
|
||||
logger.info("Sandbox session for scan %s ready and cached", scan_id)
|
||||
|
||||
@@ -3,9 +3,7 @@
|
||||
from __future__ import annotations
|
||||
|
||||
import asyncio
|
||||
import contextlib
|
||||
import json
|
||||
import logging
|
||||
import os
|
||||
import time
|
||||
import urllib.request
|
||||
@@ -28,9 +26,6 @@ if TYPE_CHECKING:
|
||||
from caido_sdk_client import Client as CaidoClient
|
||||
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
RequestPart = Literal["request", "response"]
|
||||
SortBy = Literal[
|
||||
"timestamp",
|
||||
@@ -50,18 +45,6 @@ _SITEMAP_PAGE_SIZE = 30
|
||||
_DEFAULT_CAIDO_URL = "http://127.0.0.1:48080"
|
||||
_CLIENT_CACHE: dict[str, Client] = {}
|
||||
_CLIENT_LOCK = asyncio.Lock()
|
||||
|
||||
# Substrings that mean the shared client's transport has died or is being used
|
||||
# concurrently — recoverable by rebuilding the client and retrying once.
|
||||
_CONNECTION_ERROR_MARKERS = (
|
||||
"transport is already connected",
|
||||
"connector is closed",
|
||||
"server disconnected",
|
||||
"session is closed",
|
||||
"cannot write to closing transport",
|
||||
"connection reset",
|
||||
"connection closed",
|
||||
)
|
||||
_REQ_FIELD_MAP: dict[SortBy, tuple[str, str]] = {
|
||||
"timestamp": ("req", "created_at"),
|
||||
"host": ("req", "host"),
|
||||
@@ -108,22 +91,6 @@ async def _new_client() -> Client:
|
||||
return client
|
||||
|
||||
|
||||
async def _safe_aclose(client: Client | None) -> None:
|
||||
"""Close a (possibly dead) client without letting teardown errors escape."""
|
||||
if client is None:
|
||||
return
|
||||
with contextlib.suppress(Exception):
|
||||
await client.aclose()
|
||||
|
||||
|
||||
def _is_connection_error(exc: BaseException) -> bool:
|
||||
message = str(exc).lower()
|
||||
if any(marker in message for marker in _CONNECTION_ERROR_MARKERS):
|
||||
return True
|
||||
cause = exc.__cause__ or exc.__context__
|
||||
return cause is not None and cause is not exc and _is_connection_error(cause)
|
||||
|
||||
|
||||
async def get_client() -> Client:
|
||||
"""Return the shared Caido client, creating it under a lock if needed.
|
||||
|
||||
@@ -139,73 +106,19 @@ async def get_client() -> Client:
|
||||
return client
|
||||
|
||||
|
||||
async def call_with_client[T](
|
||||
fn: Callable[[Client], Awaitable[T]], *, idempotent: bool = True
|
||||
) -> T:
|
||||
"""Run ``fn`` against the shared client, serialized and reconnect-safe.
|
||||
async def call_with_client[T](fn: Callable[[Client], Awaitable[T]]) -> T:
|
||||
"""Run ``fn`` against the shared client, serialized through ``_CLIENT_LOCK``.
|
||||
|
||||
The Caido GraphQL transport is not safe for concurrent use: two in-flight
|
||||
requests race and raise "Transport is already connected". All proxy calls
|
||||
are therefore serialized through ``_CLIENT_LOCK``. If the cached client's
|
||||
transport has since died ("Connector is closed" / "Server disconnected"),
|
||||
the stale client is closed and rebuilt so subsequent calls stop failing
|
||||
against a dead client.
|
||||
|
||||
``fn`` is only re-run automatically when ``idempotent`` is true. For
|
||||
mutations (replay, scope create/update/delete) a connection error may
|
||||
arrive *after* Caido applied the change, so we heal the client for future
|
||||
calls but re-raise instead of risking a double-apply.
|
||||
requests race and raise "Transport is already connected". Serializing every
|
||||
proxy call through the lock prevents that.
|
||||
"""
|
||||
async with _CLIENT_LOCK:
|
||||
client = _CLIENT_CACHE.get("default")
|
||||
if client is None:
|
||||
client = await _new_client()
|
||||
_CLIENT_CACHE["default"] = client
|
||||
try:
|
||||
return await fn(client)
|
||||
except Exception as exc:
|
||||
if not _is_connection_error(exc):
|
||||
raise
|
||||
new_client = await _new_client()
|
||||
_CLIENT_CACHE["default"] = new_client
|
||||
await _safe_aclose(client)
|
||||
if not idempotent:
|
||||
raise
|
||||
return await fn(new_client)
|
||||
|
||||
|
||||
class SharedCaidoClient:
|
||||
"""Serialized, reconnect-safe wrapper around one host-side Caido client.
|
||||
|
||||
Every agent in a scan shares a single instance (propagated through the
|
||||
shallow-copied run context). ``call`` serializes access — the SDK transport
|
||||
is not concurrency-safe — and, when the transport dies, rebuilds the client
|
||||
via ``reconnect`` (which preserves the Caido project) and closes the dead
|
||||
one, so a transient Caido restart no longer disables proxy tools for the
|
||||
rest of the scan.
|
||||
"""
|
||||
|
||||
def __init__(self, client: Client, reconnect: Callable[[], Awaitable[Client]]) -> None:
|
||||
self._client = client
|
||||
self._reconnect = reconnect
|
||||
self._lock = asyncio.Lock()
|
||||
|
||||
async def call[T](self, fn: Callable[[Client], Awaitable[T]], *, idempotent: bool = True) -> T:
|
||||
async with self._lock:
|
||||
try:
|
||||
return await fn(self._client)
|
||||
except Exception as exc:
|
||||
if not _is_connection_error(exc):
|
||||
raise
|
||||
dead, self._client = self._client, await self._reconnect()
|
||||
await _safe_aclose(dead)
|
||||
if not idempotent:
|
||||
raise
|
||||
return await fn(self._client)
|
||||
|
||||
async def aclose(self) -> None:
|
||||
async with self._lock:
|
||||
await _safe_aclose(self._client)
|
||||
return await fn(client)
|
||||
|
||||
|
||||
async def close_client() -> None:
|
||||
@@ -546,9 +459,7 @@ async def repeat_request(
|
||||
)
|
||||
return await replay_send_raw(client, raw=raw, connection=connection)
|
||||
|
||||
# A replay mutates server state; don't auto-retry if the transport dies
|
||||
# mid-send (the request may already have been sent).
|
||||
return await call_with_client(_run, idempotent=False)
|
||||
return await call_with_client(_run)
|
||||
|
||||
|
||||
async def scope_rules(
|
||||
@@ -569,8 +480,7 @@ async def scope_rules(
|
||||
scope_name=scope_name,
|
||||
)
|
||||
|
||||
# get/list are read-only and safe to retry; create/update/delete mutate.
|
||||
return await call_with_client(_run, idempotent=action in {"get", "list"})
|
||||
return await call_with_client(_run)
|
||||
|
||||
|
||||
async def _scope_rules_with_client(
|
||||
@@ -819,11 +729,9 @@ async def view_sitemap_entry(entry_id: str) -> dict[str, Any]:
|
||||
__all__ = [
|
||||
"RequestPart",
|
||||
"ScopeAction",
|
||||
"SharedCaidoClient",
|
||||
"SitemapDepth",
|
||||
"SortBy",
|
||||
"SortOrder",
|
||||
"call_with_client",
|
||||
"close_client",
|
||||
"get_client",
|
||||
"list_requests",
|
||||
|
||||
+48
-77
@@ -2,6 +2,7 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import asyncio
|
||||
import dataclasses
|
||||
import json
|
||||
import logging
|
||||
@@ -13,13 +14,14 @@ from typing import TYPE_CHECKING, Any, Literal
|
||||
from agents import RunContextWrapper, function_tool
|
||||
|
||||
from strix.tools.proxy import caido_api
|
||||
from strix.tools.proxy.caido_api import SharedCaidoClient
|
||||
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from collections.abc import Awaitable, Callable
|
||||
|
||||
from caido_sdk_client import Client
|
||||
|
||||
from strix.tools.proxy.caido_api import (
|
||||
@@ -29,7 +31,7 @@ if TYPE_CHECKING:
|
||||
SortOrder,
|
||||
)
|
||||
else:
|
||||
from strix.tools.proxy.caido_api import (
|
||||
from strix.tools.proxy.caido_api import ( # noqa: TC001
|
||||
RequestPart,
|
||||
SitemapDepth,
|
||||
SortBy,
|
||||
@@ -39,19 +41,21 @@ else:
|
||||
|
||||
ScopeAction = Literal["get", "list", "create", "update", "delete"]
|
||||
|
||||
# All agents in a scan share one host-side Caido client whose GraphQL transport
|
||||
# is not concurrency-safe (parallel calls raise "Transport is already
|
||||
# connected"). Serialize every host-side proxy call through this lock.
|
||||
_CAIDO_CALL_LOCK = asyncio.Lock()
|
||||
|
||||
def _ctx_proxy(ctx: RunContextWrapper) -> SharedCaidoClient | None:
|
||||
"""Return the scan-wide serialized, reconnect-safe Caido client holder.
|
||||
|
||||
All agents in a scan share one :class:`SharedCaidoClient` whose GraphQL
|
||||
transport is not concurrency-safe (parallel calls raise "Transport is
|
||||
already connected"). ``SharedCaidoClient.call`` serializes access and
|
||||
rebuilds the transport if it dies mid-scan. Returns ``None`` when no holder
|
||||
is present (e.g. standalone tool invocation outside a scan run).
|
||||
"""
|
||||
def _ctx_client(ctx: RunContextWrapper) -> Client | None:
|
||||
inner = ctx.context if isinstance(ctx.context, dict) else {}
|
||||
proxy = inner.get("caido_client")
|
||||
return proxy if isinstance(proxy, SharedCaidoClient) else None
|
||||
return inner.get("caido_client")
|
||||
|
||||
|
||||
async def _call[T](client: Client, fn: Callable[[Client], Awaitable[T]]) -> T:
|
||||
"""Run ``fn`` against the shared client, serialized under ``_CAIDO_CALL_LOCK``."""
|
||||
async with _CAIDO_CALL_LOCK:
|
||||
return await fn(client)
|
||||
|
||||
|
||||
def _to_tool_json(value: Any) -> Any:
|
||||
@@ -93,39 +97,6 @@ def _err(name: str, exc: Exception) -> str:
|
||||
)
|
||||
|
||||
|
||||
_HTTPQL_HINT = (
|
||||
"HTTPQL syntax: quote string values and leave integers unquoted; combine "
|
||||
"terms with AND / OR (there is no NOT). Numeric fields (resp.code, req.port, "
|
||||
"id, roundtrip) use eq/ne/gt/gte/lt/lte; text/byte fields (req.host, req.path, "
|
||||
"req.method, req.raw, resp.raw) use cont/ncont/eq/ne/like/nlike/regex/nregex. "
|
||||
"Example: 'resp.code.gte:200 AND resp.code.lt:300 AND req.host.cont:\"api\"'."
|
||||
)
|
||||
|
||||
|
||||
def _is_httpql_error(exc: Exception) -> bool:
|
||||
message = str(exc).lower()
|
||||
return "httpql" in message or ("filter" in message and "pars" in message)
|
||||
|
||||
|
||||
def _httpql_error(exc: Exception, httpql_filter: str | None) -> str:
|
||||
"""Return an actionable error for a rejected HTTPQL filter.
|
||||
|
||||
Preserves Caido's exact parser message and echoes the offending query so
|
||||
the agent can self-correct instead of retrying the same broken filter.
|
||||
"""
|
||||
logger.info("list_requests rejected HTTPQL filter %r: %s", httpql_filter, exc)
|
||||
return json.dumps(
|
||||
{
|
||||
"success": False,
|
||||
"error": f"Invalid HTTPQL filter: {exc}",
|
||||
"httpql_filter": httpql_filter,
|
||||
"hint": _HTTPQL_HINT,
|
||||
},
|
||||
ensure_ascii=False,
|
||||
default=str,
|
||||
)
|
||||
|
||||
|
||||
@function_tool(timeout=120)
|
||||
async def list_requests(
|
||||
ctx: RunContextWrapper,
|
||||
@@ -184,12 +155,13 @@ async def list_requests(
|
||||
sort_order: ``asc`` or ``desc``.
|
||||
scope_id: Restrict to a Caido scope (managed via ``scope_rules``).
|
||||
"""
|
||||
proxy = _ctx_proxy(ctx)
|
||||
if proxy is None:
|
||||
client = _ctx_client(ctx)
|
||||
if client is None:
|
||||
return _no_client()
|
||||
|
||||
try:
|
||||
connection = await proxy.call(
|
||||
connection = await _call(
|
||||
client,
|
||||
lambda client: caido_api.list_requests_with_client(
|
||||
client,
|
||||
httpql_filter=httpql_filter,
|
||||
@@ -198,7 +170,7 @@ async def list_requests(
|
||||
sort_by=sort_by,
|
||||
sort_order=sort_order,
|
||||
scope_id=scope_id,
|
||||
)
|
||||
),
|
||||
)
|
||||
|
||||
entries = []
|
||||
@@ -252,8 +224,6 @@ async def list_requests(
|
||||
default=str,
|
||||
)
|
||||
except Exception as exc: # noqa: BLE001
|
||||
if httpql_filter and _is_httpql_error(exc):
|
||||
return _httpql_error(exc, httpql_filter)
|
||||
return _err("list_requests", exc)
|
||||
|
||||
|
||||
@@ -291,13 +261,14 @@ async def view_request(
|
||||
page: 1-indexed page number (only when no ``search_pattern``).
|
||||
page_size: Lines per page.
|
||||
"""
|
||||
proxy = _ctx_proxy(ctx)
|
||||
if proxy is None:
|
||||
client = _ctx_client(ctx)
|
||||
if client is None:
|
||||
return _no_client()
|
||||
|
||||
try:
|
||||
result = await proxy.call(
|
||||
lambda client: caido_api.get_request_with_client(client, request_id, part=part)
|
||||
result = await _call(
|
||||
client,
|
||||
lambda client: caido_api.get_request_with_client(client, request_id, part=part),
|
||||
)
|
||||
if result is None:
|
||||
return json.dumps(
|
||||
@@ -408,8 +379,8 @@ async def repeat_request(
|
||||
- ``body`` — replace the body string entirely.
|
||||
- ``cookies`` — dict of cookies to add/update.
|
||||
"""
|
||||
proxy = _ctx_proxy(ctx)
|
||||
if proxy is None:
|
||||
client = _ctx_client(ctx)
|
||||
if client is None:
|
||||
return _no_client()
|
||||
mods = modifications or {}
|
||||
|
||||
@@ -431,9 +402,7 @@ async def repeat_request(
|
||||
return await caido_api.replay_send_raw(client, raw=raw, connection=connection)
|
||||
|
||||
try:
|
||||
# A replay mutates target state, so don't auto-retry on a mid-send
|
||||
# transport failure (the request may already have been sent).
|
||||
replay = await proxy.call(_do, idempotent=False)
|
||||
replay = await _call(client, _do)
|
||||
if replay is None:
|
||||
return json.dumps(
|
||||
{"success": False, "error": f"Request {request_id} not found"},
|
||||
@@ -492,18 +461,19 @@ async def list_sitemap(
|
||||
(recursive subtree). Only meaningful with ``parent_id``.
|
||||
page: 1-indexed page (30 entries per page).
|
||||
"""
|
||||
proxy = _ctx_proxy(ctx)
|
||||
if proxy is None:
|
||||
client = _ctx_client(ctx)
|
||||
if client is None:
|
||||
return _no_client()
|
||||
try:
|
||||
payload = await proxy.call(
|
||||
payload = await _call(
|
||||
client,
|
||||
lambda client: caido_api.list_sitemap_with_client(
|
||||
client,
|
||||
scope_id=scope_id,
|
||||
parent_id=parent_id,
|
||||
depth=depth,
|
||||
page=page,
|
||||
)
|
||||
),
|
||||
)
|
||||
return json.dumps(payload, ensure_ascii=False, default=str)
|
||||
except Exception as exc: # noqa: BLE001
|
||||
@@ -525,12 +495,13 @@ async def view_sitemap_entry(
|
||||
Args:
|
||||
entry_id: ID from ``list_sitemap`` (or any nested entry).
|
||||
"""
|
||||
proxy = _ctx_proxy(ctx)
|
||||
if proxy is None:
|
||||
client = _ctx_client(ctx)
|
||||
if client is None:
|
||||
return _no_client()
|
||||
try:
|
||||
payload = await proxy.call(
|
||||
lambda client: caido_api.view_sitemap_entry_with_client(client, entry_id)
|
||||
payload = await _call(
|
||||
client,
|
||||
lambda client: caido_api.view_sitemap_entry_with_client(client, entry_id),
|
||||
)
|
||||
return json.dumps(payload, ensure_ascii=False, default=str)
|
||||
except Exception as exc: # noqa: BLE001
|
||||
@@ -583,13 +554,13 @@ async def scope_rules(
|
||||
scope_id: Required for ``get`` / ``update`` / ``delete``.
|
||||
scope_name: Required for ``create`` / ``update``.
|
||||
"""
|
||||
proxy = _ctx_proxy(ctx)
|
||||
if proxy is None:
|
||||
client = _ctx_client(ctx)
|
||||
if client is None:
|
||||
return _no_client()
|
||||
|
||||
try:
|
||||
if action == "list":
|
||||
scopes = await proxy.call(caido_api.scope_list)
|
||||
scopes = await _call(client, caido_api.scope_list)
|
||||
return json.dumps(
|
||||
{"success": True, "scopes": [_to_tool_json(s) for s in scopes]},
|
||||
ensure_ascii=False,
|
||||
@@ -602,7 +573,7 @@ async def scope_rules(
|
||||
ensure_ascii=False,
|
||||
default=str,
|
||||
)
|
||||
scope = await proxy.call(lambda client: caido_api.scope_get(client, scope_id))
|
||||
scope = await _call(client, lambda client: caido_api.scope_get(client, scope_id))
|
||||
return json.dumps(
|
||||
{"success": True, "scope": _to_tool_json(scope)},
|
||||
ensure_ascii=False,
|
||||
@@ -615,11 +586,11 @@ async def scope_rules(
|
||||
ensure_ascii=False,
|
||||
default=str,
|
||||
)
|
||||
scope = await proxy.call(
|
||||
scope = await _call(
|
||||
client,
|
||||
lambda client: caido_api.scope_create(
|
||||
client, name=scope_name, allowlist=allowlist, denylist=denylist
|
||||
),
|
||||
idempotent=False,
|
||||
)
|
||||
return json.dumps(
|
||||
{"success": True, "scope": _to_tool_json(scope)},
|
||||
@@ -636,11 +607,11 @@ async def scope_rules(
|
||||
ensure_ascii=False,
|
||||
default=str,
|
||||
)
|
||||
scope = await proxy.call(
|
||||
scope = await _call(
|
||||
client,
|
||||
lambda client: caido_api.scope_update(
|
||||
client, scope_id, name=scope_name, allowlist=allowlist, denylist=denylist
|
||||
),
|
||||
idempotent=False,
|
||||
)
|
||||
return json.dumps(
|
||||
{"success": True, "scope": _to_tool_json(scope)},
|
||||
@@ -653,7 +624,7 @@ async def scope_rules(
|
||||
ensure_ascii=False,
|
||||
default=str,
|
||||
)
|
||||
await proxy.call(lambda client: caido_api.scope_delete(client, scope_id), idempotent=False)
|
||||
await _call(client, lambda client: caido_api.scope_delete(client, scope_id))
|
||||
return json.dumps(
|
||||
{
|
||||
"success": True,
|
||||
|
||||
+17
-182
@@ -1,20 +1,19 @@
|
||||
"""Tests for the shared Caido client lifecycle and proxy error handling.
|
||||
"""Tests for the shared Caido client lifecycle and proxy call serialization.
|
||||
|
||||
Covers the concurrency/reconnect guarantees of ``caido_api.call_with_client``
|
||||
(the sandbox-imported path) and ``caido_api.SharedCaidoClient`` (the host-side
|
||||
holder), plus the actionable HTTPQL errors in ``proxy.tools``.
|
||||
Covers the caching + serialization guarantees of ``caido_api.call_with_client``
|
||||
(the sandbox-imported path) and ``proxy.tools._call`` (the host-side path). The
|
||||
Caido GraphQL transport is not concurrency-safe, so both paths must run one
|
||||
call at a time against the shared client.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import asyncio
|
||||
import json
|
||||
from typing import TYPE_CHECKING, Any, cast
|
||||
|
||||
import pytest
|
||||
|
||||
from strix.tools.proxy import caido_api, tools
|
||||
from strix.tools.proxy.caido_api import SharedCaidoClient
|
||||
|
||||
|
||||
if TYPE_CHECKING:
|
||||
@@ -91,83 +90,15 @@ async def test_failed_init_does_not_poison_cache(monkeypatch: pytest.MonkeyPatch
|
||||
assert "default" not in caido_api._CLIENT_CACHE
|
||||
|
||||
|
||||
async def test_call_with_client_reconnects_and_closes_dead_transport(
|
||||
monkeypatch: pytest.MonkeyPatch,
|
||||
) -> None:
|
||||
dead = _FakeClient("dead")
|
||||
fresh = _FakeClient("fresh")
|
||||
caido_api._CLIENT_CACHE["default"] = cast("Any", dead)
|
||||
|
||||
new_calls = {"n": 0}
|
||||
|
||||
async def _new() -> Any:
|
||||
new_calls["n"] += 1
|
||||
return fresh
|
||||
|
||||
monkeypatch.setattr(caido_api, "_new_client", _new)
|
||||
|
||||
attempts: list[Any] = []
|
||||
|
||||
async def fn(client: Any) -> str:
|
||||
attempts.append(client)
|
||||
if len(attempts) == 1:
|
||||
raise RuntimeError("Transport is already connected")
|
||||
return "ok"
|
||||
|
||||
assert await caido_api.call_with_client(fn) == "ok"
|
||||
assert attempts == [dead, fresh]
|
||||
assert new_calls["n"] == 1
|
||||
assert caido_api._CLIENT_CACHE["default"] is fresh
|
||||
assert dead.closed is True # stale transport is not leaked
|
||||
|
||||
|
||||
async def test_call_with_client_non_idempotent_rebuilds_but_reraises(
|
||||
monkeypatch: pytest.MonkeyPatch,
|
||||
) -> None:
|
||||
dead = _FakeClient("dead")
|
||||
fresh = _FakeClient("fresh")
|
||||
caido_api._CLIENT_CACHE["default"] = cast("Any", dead)
|
||||
|
||||
async def _new() -> Any:
|
||||
return fresh
|
||||
|
||||
monkeypatch.setattr(caido_api, "_new_client", _new)
|
||||
|
||||
calls = {"n": 0}
|
||||
|
||||
async def fn(_client: Any) -> str:
|
||||
calls["n"] += 1
|
||||
raise RuntimeError("Server disconnected")
|
||||
|
||||
# A mutation must not be auto-retried (it may already have applied), but the
|
||||
# dead client is still healed so later calls succeed.
|
||||
with pytest.raises(RuntimeError, match="Server disconnected"):
|
||||
await caido_api.call_with_client(fn, idempotent=False)
|
||||
assert calls["n"] == 1
|
||||
assert caido_api._CLIENT_CACHE["default"] is fresh
|
||||
assert dead.closed is True
|
||||
|
||||
|
||||
async def test_call_with_client_does_not_retry_application_errors(
|
||||
monkeypatch: pytest.MonkeyPatch,
|
||||
) -> None:
|
||||
async def test_call_with_client_propagates_errors() -> None:
|
||||
cached = _FakeClient("cached")
|
||||
caido_api._CLIENT_CACHE["default"] = cast("Any", cached)
|
||||
|
||||
async def _new() -> Any:
|
||||
raise AssertionError("deterministic errors must not trigger a reconnect")
|
||||
|
||||
monkeypatch.setattr(caido_api, "_new_client", _new)
|
||||
|
||||
calls = {"n": 0}
|
||||
|
||||
async def fn(_client: Any) -> str:
|
||||
calls["n"] += 1
|
||||
raise ValueError("Invalid HTTPQL filter")
|
||||
|
||||
with pytest.raises(ValueError, match="Invalid HTTPQL"):
|
||||
await caido_api.call_with_client(fn)
|
||||
assert calls["n"] == 1
|
||||
assert caido_api._CLIENT_CACHE["default"] is cached
|
||||
|
||||
|
||||
@@ -177,7 +108,7 @@ async def test_call_with_client_serializes_concurrent_calls(
|
||||
caido_api._CLIENT_CACHE["default"] = cast("Any", _FakeClient("shared"))
|
||||
|
||||
async def _new() -> Any:
|
||||
raise AssertionError("no reconnect expected")
|
||||
raise AssertionError("no new client expected")
|
||||
|
||||
monkeypatch.setattr(caido_api, "_new_client", _new)
|
||||
|
||||
@@ -194,61 +125,8 @@ async def test_call_with_client_serializes_concurrent_calls(
|
||||
assert state["max"] == 1
|
||||
|
||||
|
||||
async def test_shared_client_reconnects_and_closes_dead_transport() -> None:
|
||||
dead = _FakeClient("dead")
|
||||
fresh = _FakeClient("fresh")
|
||||
|
||||
async def _reconnect() -> Any:
|
||||
return fresh
|
||||
|
||||
holder = SharedCaidoClient(cast("Any", dead), _reconnect)
|
||||
|
||||
attempts: list[Any] = []
|
||||
|
||||
async def fn(client: Any) -> str:
|
||||
attempts.append(client)
|
||||
if len(attempts) == 1:
|
||||
raise RuntimeError("Connector is closed")
|
||||
return "ok"
|
||||
|
||||
assert await holder.call(fn) == "ok"
|
||||
assert attempts == [dead, fresh]
|
||||
assert dead.closed is True
|
||||
|
||||
|
||||
async def test_shared_client_non_idempotent_rebuilds_but_reraises() -> None:
|
||||
dead = _FakeClient("dead")
|
||||
fresh = _FakeClient("fresh")
|
||||
|
||||
async def _reconnect() -> Any:
|
||||
return fresh
|
||||
|
||||
holder = SharedCaidoClient(cast("Any", dead), _reconnect)
|
||||
|
||||
calls = {"n": 0}
|
||||
|
||||
async def fn(_client: Any) -> str:
|
||||
calls["n"] += 1
|
||||
raise RuntimeError("Server disconnected")
|
||||
|
||||
with pytest.raises(RuntimeError, match="Server disconnected"):
|
||||
await holder.call(fn, idempotent=False)
|
||||
assert calls["n"] == 1
|
||||
assert dead.closed is True
|
||||
# The healthy client remains for the next call.
|
||||
assert await holder.call(lambda _c: _ok()) == "ok"
|
||||
|
||||
|
||||
async def _ok() -> str:
|
||||
return "ok"
|
||||
|
||||
|
||||
async def test_shared_client_serializes_concurrent_calls() -> None:
|
||||
async def _reconnect() -> Any:
|
||||
raise AssertionError("no reconnect expected")
|
||||
|
||||
holder = SharedCaidoClient(cast("Any", _FakeClient("shared")), _reconnect)
|
||||
|
||||
async def test_host_call_serializes_concurrent_calls() -> None:
|
||||
client = _FakeClient("host")
|
||||
state = {"active": 0, "max": 0}
|
||||
|
||||
async def fn(_client: Any) -> str:
|
||||
@@ -258,64 +136,21 @@ async def test_shared_client_serializes_concurrent_calls() -> None:
|
||||
state["active"] -= 1
|
||||
return "ok"
|
||||
|
||||
await asyncio.gather(*(holder.call(fn) for _ in range(6)))
|
||||
await asyncio.gather(*(tools._call(cast("Any", client), fn) for _ in range(6)))
|
||||
assert state["max"] == 1
|
||||
|
||||
|
||||
async def test_shared_client_passes_through_application_errors() -> None:
|
||||
async def _reconnect() -> Any:
|
||||
raise AssertionError("deterministic errors must not trigger a reconnect")
|
||||
|
||||
holder = SharedCaidoClient(cast("Any", _FakeClient("c")), _reconnect)
|
||||
|
||||
async def fn(_client: Any) -> str:
|
||||
raise ValueError("Invalid HTTPQL filter")
|
||||
|
||||
with pytest.raises(ValueError, match="Invalid HTTPQL"):
|
||||
await holder.call(fn)
|
||||
|
||||
|
||||
def test_is_connection_error_matches_markers_and_causes() -> None:
|
||||
assert caido_api._is_connection_error(RuntimeError("Transport is already connected"))
|
||||
assert caido_api._is_connection_error(RuntimeError("Connector is closed"))
|
||||
assert caido_api._is_connection_error(RuntimeError("Server disconnected"))
|
||||
assert not caido_api._is_connection_error(ValueError("Invalid HTTPQL filter"))
|
||||
|
||||
nested = RuntimeError("wrapper")
|
||||
nested.__cause__ = RuntimeError("connection reset by peer")
|
||||
assert caido_api._is_connection_error(nested)
|
||||
|
||||
|
||||
class _Ctx:
|
||||
def __init__(self, context: Any) -> None:
|
||||
self.context = context
|
||||
|
||||
|
||||
def test_ctx_proxy_returns_holder_when_present() -> None:
|
||||
async def _reconnect() -> Any:
|
||||
raise AssertionError("unused")
|
||||
|
||||
holder = SharedCaidoClient(cast("Any", _FakeClient("c")), _reconnect)
|
||||
got = tools._ctx_proxy(cast("Any", _Ctx({"caido_client": holder})))
|
||||
assert got is holder
|
||||
def test_ctx_client_returns_client_when_present() -> None:
|
||||
client = _FakeClient("host")
|
||||
got = tools._ctx_client(cast("Any", _Ctx({"caido_client": client})))
|
||||
assert got is client
|
||||
|
||||
|
||||
def test_ctx_proxy_returns_none_without_holder() -> None:
|
||||
assert tools._ctx_proxy(cast("Any", _Ctx({}))) is None
|
||||
assert tools._ctx_proxy(cast("Any", _Ctx(None))) is None
|
||||
assert tools._ctx_proxy(cast("Any", _Ctx({"caido_client": object()}))) is None
|
||||
|
||||
|
||||
def test_is_httpql_error_detection() -> None:
|
||||
assert tools._is_httpql_error(RuntimeError("HTTPQL parse error at column 4"))
|
||||
assert tools._is_httpql_error(RuntimeError("failed to parse filter"))
|
||||
assert not tools._is_httpql_error(RuntimeError("Transport is already connected"))
|
||||
|
||||
|
||||
def test_httpql_error_preserves_message_and_query() -> None:
|
||||
exc = RuntimeError("HTTPQL parse error: unexpected token at column 12")
|
||||
payload = json.loads(tools._httpql_error(exc, 'resp.code.eq:"200"'))
|
||||
assert payload["success"] is False
|
||||
assert "unexpected token at column 12" in payload["error"]
|
||||
assert payload["httpql_filter"] == 'resp.code.eq:"200"'
|
||||
assert "AND / OR" in payload["hint"]
|
||||
def test_ctx_client_returns_none_without_client() -> None:
|
||||
assert tools._ctx_client(cast("Any", _Ctx({}))) is None
|
||||
assert tools._ctx_client(cast("Any", _Ctx(None))) is None
|
||||
|
||||
Reference in New Issue
Block a user