mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-31 19:16:29 +00:00
Second, deeper pass over tools/gateway/hermes_cli plus first pass over the trees wave 1 missed (acp, acp_adapter, skills, computer_use, docker, dashboard, conformance, monitoring, secret_sources, hermes_state, providers). Same rubric as wave 1 (AGENTS.md test policy); security, alternation/caching invariants, issue-number regressions, and E2E kept. Real test-quality fixes found and rooted out along the way: - tests/tools/test_command_guards.py made real auxiliary-LLM HTTPS calls (DEFAULT_CONFIG smart-approval leaked in) — pinned approval mode=manual via autouse fixture: 17.4s → 0.4s. - test_model_switch_custom_providers.py / test_user_providers_model_switch.py silently probed live provider catalogs (~2s/test) — stubbed cached_provider_model_ids/provider_model_ids/fetch_api_models. - test_telegram_noise_filter.py: 15-platform copy-paste matrix over shared gateway.run logic → 3 representative platforms (55s → 3.9s). - test_gateway_shutdown.py: stop()'s 5s interrupt-deadline loop spun on MagicMock agents — interrupt.side_effect now clears _running_agents (22s → 1.0s). - test_gateway_inactivity_timeout.py poll-harness timings shrunk 3-5x (24s → 1.1s); test_mcp_stability.py backoff/SIGTERM-grace sleeps patched (15.4s → 2.5s); test_async_delegation.py negative-drain wait 5s → 0.5s. - test_telegram_init_deadline.py: loop-block margin restored to 1.0s with rationale comment — the watchdog-dump assertion needs the loop blocked well past deadline+grace under parallel load (flaked once in the 40-worker verification run at a 0.2s margin). Verification: full hermetic suite via scripts/run_tests.sh — 2,438 files, 21,718 tests passed, 0 failed, 293.9s wall. Suite totals vs original baseline: 46,820 → 19,757 test functions (−57.8%), wall 583.5s → 293.9s (−50%), subprocess CPU 13,564s → 11,623s.
75 lines
3.1 KiB
Python
75 lines
3.1 KiB
Python
"""Tests for delegate_tool toolset scoping.
|
|
|
|
Verifies that subagents cannot gain tools that the parent does not have.
|
|
The LLM controls the `toolsets` parameter — without intersection with the
|
|
parent's enabled_toolsets, it can escalate privileges by requesting
|
|
arbitrary toolsets.
|
|
"""
|
|
|
|
from types import SimpleNamespace
|
|
|
|
from tools.delegate_tool import _strip_blocked_tools, _emit_parent_console
|
|
|
|
|
|
class TestToolsetIntersection:
|
|
"""Subagent toolsets must be a subset of parent's enabled_toolsets."""
|
|
|
|
def test_requested_toolsets_intersected_with_parent(self):
|
|
"""LLM requests toolsets parent doesn't have — extras are dropped."""
|
|
parent = SimpleNamespace(enabled_toolsets=["terminal", "file"])
|
|
|
|
# Simulate the intersection logic from _build_child_agent
|
|
parent_toolsets = set(parent.enabled_toolsets)
|
|
requested = ["terminal", "file", "web", "browser", "rl"]
|
|
scoped = [t for t in requested if t in parent_toolsets]
|
|
|
|
assert sorted(scoped) == ["file", "terminal"]
|
|
assert "web" not in scoped
|
|
assert "browser" not in scoped
|
|
assert "rl" not in scoped
|
|
|
|
|
|
def test_strip_blocked_removes_delegation(self):
|
|
"""Blocked toolsets (delegation, clarify, etc.) are always removed."""
|
|
child = _strip_blocked_tools(["terminal", "delegation", "clarify", "memory"])
|
|
assert "delegation" not in child
|
|
assert "clarify" not in child
|
|
assert "memory" not in child
|
|
assert "terminal" in child
|
|
|
|
def test_empty_intersection_yields_empty_toolsets(self):
|
|
"""If parent has no overlap with requested, child gets nothing extra."""
|
|
parent = SimpleNamespace(enabled_toolsets=["terminal"])
|
|
|
|
parent_toolsets = set(parent.enabled_toolsets)
|
|
requested = ["web", "browser"]
|
|
scoped = [t for t in requested if t in parent_toolsets]
|
|
|
|
assert scoped == []
|
|
|
|
|
|
class TestEmitParentConsole:
|
|
"""Progress lines (e.g. ``✓ [N/M] …``) must route through the parent's
|
|
configured ``_safe_print`` in headless stdio hosts (ACP, gateway) so
|
|
they don't land on stdout and corrupt JSON-RPC frames. Regression for a
|
|
bug where delegate_task completion lines pushed to stdout caused
|
|
``Failed to parse JSON message: ✓ [3/3] …`` errors in the ACP adapter."""
|
|
|
|
def test_routes_through_parent_safe_print_when_available(self, capsys):
|
|
captured_lines = []
|
|
parent = SimpleNamespace(_safe_print=lambda line: captured_lines.append(line))
|
|
|
|
_emit_parent_console(parent, " ✓ [1/3] Research done (11.55s)")
|
|
|
|
assert captured_lines == [" ✓ [1/3] Research done (11.55s)"]
|
|
stdout_stderr = capsys.readouterr()
|
|
assert stdout_stderr.out == ""
|
|
assert stdout_stderr.err == ""
|
|
|
|
|
|
def test_non_callable_safe_print_is_ignored(self, capsys):
|
|
"""Defensive: if _safe_print is set but not callable, fall back."""
|
|
parent = SimpleNamespace(_safe_print="not-a-function")
|
|
_emit_parent_console(parent, " ✓ [3/3] non-callable guard")
|
|
captured = capsys.readouterr()
|
|
assert "non-callable guard" in captured.out
|