test: close critical coverage gaps and harden PR review fixes

Critical bug fixes from PR review agents:
- client.py: eager asyncio.Lock init, Final[frozenset] for _SENSITIVE_KEYS,
  explicit 429 ToolError after retries exhausted, removed lazy _get_client_lock()
  and _RateLimiter._get_lock() patterns
- exceptions.py: use builtin TimeoutError (UP041), explicit handler before broad
  except so asyncio timeouts get descriptive messages
- docker.py: add update_all to DESTRUCTIVE_ACTIONS (was missing), remove dead
  _MUTATION_ACTIONS constant
- manager.py: _cap_log_content returns new dict (immutable), lock write to
  resource_data, clean dead task from active_subscriptions after loop exits
- diagnostics.py: fix inaccurate comment about semicolon injection guard
- health.py: narrow except ValueError in _safe_display_url, fix TODO comment

New test coverage (98 tests added, 529 → 598 passing):
- test_subscription_validation.py: 27 tests for _validate_subscription_query
  (security-critical allow-list, forbidden keyword guards, word-boundary test)
- test_subscription_manager.py: 12 tests for _cap_log_content
  (immutability, truncation, nesting, passthrough)
- test_client.py: +57 tests — _RateLimiter (token math, refill, sleep-on-empty),
  _QueryCache (TTL, invalidation, is_cacheable), 429 retry loop (1/2/3 failures)
- test_health.py: +10 tests for _safe_display_url (credential strip, port,
  path/query removal, malformed IPv6 → <unparseable>)
- test_notifications.py: +7 importance enum and field length validation tests
- test_rclone.py: +7 _validate_config_data security guard tests
- test_storage.py: +15 (tail_lines bounds, format_kb, safe_get)
- test_docker.py: update_all now requires confirm=True + new guard test
- test_destructive_guards.py: update audit to include update_all

Co-authored-by: Claude <noreply@anthropic.com>
This commit is contained in:
Jacob Magar
2026-02-18 01:28:40 -05:00
parent 316193c04b
commit f76e676fd4
15 changed files with 867 additions and 56 deletions

View File

@@ -6,7 +6,7 @@ throughout the application, with proper integration to FastMCP's error system.
import contextlib
import logging
from collections.abc import Generator
from collections.abc import Iterator
from fastmcp.exceptions import ToolError as FastMCPToolError
@@ -28,11 +28,12 @@ def tool_error_handler(
tool_name: str,
action: str,
logger: logging.Logger,
) -> Generator[None]:
) -> Iterator[None]:
"""Context manager that standardizes tool error handling.
Re-raises ToolError as-is. Catches all other exceptions, logs them
with full traceback, and wraps them in ToolError with a descriptive message.
Re-raises ToolError as-is. Gives TimeoutError a descriptive message.
Catches all other exceptions, logs them with full traceback, and wraps them
in ToolError with a descriptive message.
Args:
tool_name: The tool name for error messages (e.g., "docker", "vm").
@@ -43,6 +44,14 @@ def tool_error_handler(
yield
except ToolError:
raise
except TimeoutError as e:
logger.error(
f"Timeout in unraid_{tool_name} action={action}: request exceeded time limit",
exc_info=True,
)
raise ToolError(
f"Request timed out executing {tool_name}/{action}. The Unraid API did not respond in time."
) from e
except Exception as e:
logger.error(f"Error in unraid_{tool_name} action={action}: {e}", exc_info=True)
raise ToolError(f"Failed to execute {tool_name}/{action}: {e!s}") from e