mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-30 19:09:28 +00:00
6b81590c55
Systematic prune per AGENTS.md test policy, one pass over every major test tree (gateway, hermes_cli, tools, agent, run_agent, plugins, cli, cron, tui_gateway, honcho/openviking, root-level): - DELETE: source-reading tests (read_text/getsource on prod files), change-detector tests (exact catalog counts, model-name snapshots, config version literals), mock-echo tests (assert a mock returns what it was told), assertion-free/trivial tests, near-duplicate parametrizations (boundaries + one representative kept), async/sync twin duplicates, cosmetic within-file variations. - KEEP (mandatory): security/redaction/approval guards, message-role alternation invariants, prompt-caching/deterministic-call-id invariants, issue-number regression tests (deduped), E2E tests. - 6 test files deleted outright (script-style/no-assert or fully redundant); conftest.py, fakes/, fixtures/ untouched. - tests/acp/conftest.py added: autouse fixture stubs the live models.dev/GitHub/Copilot/Anthropic inventory fetches that ACP server tests performed on every session create — test_server.py 147s → 3.4s, and the tests are now genuinely hermetic. - Sleep-based slowness shrunk where safe (codex_ttfb_watchdog, compression_concurrent_fork, etc.); no wall-clock assertion tightened. Verification: full hermetic suite via scripts/run_tests.sh — 2439 files, 31,130 tests passed, 0 failed, 0 flaky retries, 315s wall (baseline: 583s wall, 13,564s subprocess CPU).
271 lines
9.2 KiB
Python
271 lines
9.2 KiB
Python
"""Tests for toolsets.py — toolset resolution, validation, and composition."""
|
|
|
|
from tools.registry import ToolRegistry
|
|
from toolsets import (
|
|
TOOLSETS,
|
|
get_toolset,
|
|
resolve_toolset,
|
|
resolve_multiple_toolsets,
|
|
get_all_toolsets,
|
|
validate_toolset,
|
|
create_custom_toolset,
|
|
get_toolset_info,
|
|
)
|
|
|
|
|
|
def _dummy_handler(args, **kwargs):
|
|
return "{}"
|
|
|
|
|
|
def _make_schema(name: str, description: str = "test tool"):
|
|
return {
|
|
"name": name,
|
|
"description": description,
|
|
"parameters": {"type": "object", "properties": {}},
|
|
}
|
|
|
|
|
|
class TestGetToolset:
|
|
def test_known_toolset(self):
|
|
ts = get_toolset("web")
|
|
assert ts is not None
|
|
assert "web_search" in ts["tools"]
|
|
|
|
def test_x_search_toolset_marks_read_only_and_points_to_xurl(self):
|
|
ts = get_toolset("x_search")
|
|
assert ts is not None
|
|
assert ts["tools"] == ["x_search"]
|
|
description = ts["description"].lower()
|
|
assert "read-only" in description
|
|
assert "xurl" in description
|
|
assert "authenticated" in description
|
|
|
|
def test_merges_registry_tools_into_builtin_toolset(self, monkeypatch):
|
|
reg = ToolRegistry()
|
|
reg.register(
|
|
name="web_search_plus",
|
|
toolset="web",
|
|
schema=_make_schema("web_search_plus", "Plugin web search"),
|
|
handler=_dummy_handler,
|
|
)
|
|
|
|
monkeypatch.setattr("tools.registry.registry", reg)
|
|
|
|
ts = get_toolset("web")
|
|
assert ts is not None
|
|
assert set(ts["tools"]) == {"web_search", "web_extract", "web_search_plus"}
|
|
|
|
|
|
|
|
class TestResolveToolset:
|
|
def test_leaf_toolset(self):
|
|
tools = resolve_toolset("web")
|
|
assert set(tools) == {"web_search", "web_extract"}
|
|
|
|
def test_composite_toolset(self):
|
|
tools = resolve_toolset("debugging")
|
|
assert "terminal" in tools
|
|
assert "web_search" in tools
|
|
assert "web_extract" in tools
|
|
|
|
def test_cycle_detection(self):
|
|
# Create a cycle: A includes B, B includes A
|
|
TOOLSETS["_cycle_a"] = {"description": "test", "tools": ["t1"], "includes": ["_cycle_b"]}
|
|
TOOLSETS["_cycle_b"] = {"description": "test", "tools": ["t2"], "includes": ["_cycle_a"]}
|
|
try:
|
|
tools = resolve_toolset("_cycle_a")
|
|
# Should not infinite loop — cycle is detected
|
|
assert "t1" in tools
|
|
assert "t2" in tools
|
|
finally:
|
|
del TOOLSETS["_cycle_a"]
|
|
del TOOLSETS["_cycle_b"]
|
|
|
|
|
|
def test_plugin_toolset_uses_registry_snapshot(self, monkeypatch):
|
|
reg = ToolRegistry()
|
|
reg.register(
|
|
name="plugin_b",
|
|
toolset="plugin_example",
|
|
schema=_make_schema("plugin_b", "B"),
|
|
handler=_dummy_handler,
|
|
)
|
|
reg.register(
|
|
name="plugin_a",
|
|
toolset="plugin_example",
|
|
schema=_make_schema("plugin_a", "A"),
|
|
handler=_dummy_handler,
|
|
)
|
|
|
|
monkeypatch.setattr("tools.registry.registry", reg)
|
|
|
|
assert resolve_toolset("plugin_example") == ["plugin_a", "plugin_b"]
|
|
|
|
|
|
|
|
|
|
class TestResolveMultipleToolsets:
|
|
def test_combines_and_deduplicates(self):
|
|
tools = resolve_multiple_toolsets(["web", "terminal"])
|
|
assert "web_search" in tools
|
|
assert "web_extract" in tools
|
|
assert "terminal" in tools
|
|
# No duplicates
|
|
assert len(tools) == len(set(tools))
|
|
|
|
|
|
|
|
class TestValidateToolset:
|
|
def test_valid(self):
|
|
assert validate_toolset("web") is True
|
|
assert validate_toolset("terminal") is True
|
|
|
|
|
|
def test_invalid(self):
|
|
assert validate_toolset("nonexistent") is False
|
|
|
|
def test_mcp_alias_uses_live_registry(self, monkeypatch):
|
|
reg = ToolRegistry()
|
|
reg.register(
|
|
name="mcp__dynserver__ping",
|
|
toolset="mcp-dynserver",
|
|
schema=_make_schema("mcp__dynserver__ping", "Ping"),
|
|
handler=_dummy_handler,
|
|
)
|
|
reg.register_toolset_alias("dynserver", "mcp-dynserver")
|
|
|
|
monkeypatch.setattr("tools.registry.registry", reg)
|
|
|
|
assert validate_toolset("dynserver") is True
|
|
assert validate_toolset("mcp-dynserver") is True
|
|
assert "mcp__dynserver__ping" in resolve_toolset("dynserver")
|
|
|
|
|
|
class TestGetToolsetInfo:
|
|
def test_leaf(self):
|
|
info = get_toolset_info("web")
|
|
assert info["name"] == "web"
|
|
assert info["is_composite"] is False
|
|
assert info["tool_count"] == 2
|
|
|
|
def test_composite(self):
|
|
info = get_toolset_info("debugging")
|
|
assert info["is_composite"] is True
|
|
assert info["tool_count"] > len(info["direct_tools"])
|
|
|
|
|
|
|
|
class TestCreateCustomToolset:
|
|
def test_runtime_creation(self):
|
|
create_custom_toolset(
|
|
name="_test_custom",
|
|
description="Test toolset",
|
|
tools=["web_search"],
|
|
includes=["terminal"],
|
|
)
|
|
try:
|
|
tools = resolve_toolset("_test_custom")
|
|
assert "web_search" in tools
|
|
assert "terminal" in tools
|
|
assert validate_toolset("_test_custom") is True
|
|
finally:
|
|
del TOOLSETS["_test_custom"]
|
|
|
|
|
|
class TestRegistryOwnedToolsets:
|
|
def test_registry_membership_is_live(self, monkeypatch):
|
|
reg = ToolRegistry()
|
|
reg.register(
|
|
name="test_live_toolset_tool",
|
|
toolset="test-live-toolset",
|
|
schema=_make_schema("test_live_toolset_tool", "Live"),
|
|
handler=_dummy_handler,
|
|
)
|
|
|
|
monkeypatch.setattr("tools.registry.registry", reg)
|
|
|
|
assert validate_toolset("test-live-toolset") is True
|
|
assert get_toolset("test-live-toolset")["tools"] == ["test_live_toolset_tool"]
|
|
assert resolve_toolset("test-live-toolset") == ["test_live_toolset_tool"]
|
|
|
|
|
|
class TestToolsetConsistency:
|
|
"""Verify structural integrity of the built-in TOOLSETS dict."""
|
|
|
|
def test_all_toolsets_have_required_keys(self):
|
|
for name, ts in TOOLSETS.items():
|
|
assert "description" in ts, f"{name} missing description"
|
|
assert "tools" in ts, f"{name} missing tools"
|
|
assert "includes" in ts, f"{name} missing includes"
|
|
|
|
|
|
def test_hermes_platforms_share_core_tools(self):
|
|
"""All hermes-* platform toolsets share the same core tools.
|
|
|
|
Platform-specific additions (e.g. ``discord`` / ``discord_admin``
|
|
on hermes-discord, gated on DISCORD_BOT_TOKEN) are allowed on top —
|
|
the invariant is that the core set is identical across platforms.
|
|
"""
|
|
platforms = ["hermes-cli", "hermes-telegram", "hermes-discord", "hermes-whatsapp", "hermes-slack", "hermes-signal", "hermes-homeassistant"]
|
|
tool_sets = [set(TOOLSETS[p]["tools"]) for p in platforms]
|
|
# All platforms must contain the shared core; platform-specific
|
|
# extras are OK (subset check, not equality).
|
|
core = set.intersection(*tool_sets)
|
|
for name, ts in zip(platforms, tool_sets):
|
|
assert core.issubset(ts), f"{name} is missing core tools: {core - ts}"
|
|
# Sanity: the shared core must be non-trivial (i.e. we didn't
|
|
# silently let a platform diverge so far that nothing is shared).
|
|
assert len(core) > 20, f"Suspiciously small shared core: {len(core)} tools"
|
|
|
|
|
|
class TestPluginToolsets:
|
|
def test_get_all_toolsets_includes_plugin_toolset(self, monkeypatch):
|
|
reg = ToolRegistry()
|
|
reg.register(
|
|
name="plugin_tool",
|
|
toolset="plugin_bundle",
|
|
schema=_make_schema("plugin_tool", "Plugin tool"),
|
|
handler=_dummy_handler,
|
|
)
|
|
|
|
monkeypatch.setattr("tools.registry.registry", reg)
|
|
|
|
all_toolsets = get_all_toolsets()
|
|
assert "plugin_bundle" in all_toolsets
|
|
assert all_toolsets["plugin_bundle"]["tools"] == ["plugin_tool"]
|
|
|
|
|
|
class TestDefaultPlatformWebSearchCoverage:
|
|
def test_hermes_whatsapp_toolset_includes_web_search(self):
|
|
assert "web_search" in resolve_toolset("hermes-whatsapp")
|
|
|
|
|
|
|
|
class TestResolveToolsetIncludeRegistry:
|
|
"""include_registry flag exposes the static (pre-registry-merge) view used
|
|
by platform reverse-mapping. Regression harness for issue #49622."""
|
|
|
|
def test_include_registry_false_excludes_registry_tools(self):
|
|
from tools.registry import discover_builtin_tools
|
|
discover_builtin_tools() # registers read_terminal into 'terminal'
|
|
|
|
merged = set(resolve_toolset("terminal"))
|
|
static = set(resolve_toolset("terminal", include_registry=False))
|
|
|
|
assert static == {"terminal", "process"}, static
|
|
# read_terminal is registered into 'terminal' but is desktop-only and
|
|
# not part of the static definition — it must only appear in the merged view.
|
|
assert "read_terminal" in merged
|
|
assert "read_terminal" not in static
|
|
|
|
|
|
def test_static_view_threads_through_includes(self):
|
|
# 'debugging' has direct tools [terminal, process] and includes [web, file]
|
|
static = set(resolve_toolset("debugging", include_registry=False))
|
|
assert {"terminal", "process"} <= static
|
|
assert "web_search" in static
|
|
assert "read_file" in static
|
|
|
|
|
|
def test_registry_only_toolset_static_view_is_empty(self):
|
|
assert resolve_toolset("__definitely_not_a_real_toolset__", include_registry=False) == []
|