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.
95 lines
3.5 KiB
Python
95 lines
3.5 KiB
Python
"""Conformance vector generator tests (Phase 5 oracle workstream).
|
|
|
|
Behavior contracts, not change-detectors: determinism, corpus invariants,
|
|
oracle-import health (the generator calls format_message UNBOUND — these
|
|
tests fail loudly if a refactor gives those renderers instance state), and
|
|
vector-file shape. Native outputs themselves are NOT snapshotted here — the
|
|
connector's conformance runner is the consumer that asserts them.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
import re
|
|
import sys
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
|
|
REPO_ROOT = Path(__file__).resolve().parents[2]
|
|
sys.path.insert(0, str(REPO_ROOT / "scripts"))
|
|
|
|
from generate_conformance_vectors import ( # noqa: E402
|
|
ADVERSARIAL,
|
|
GRID,
|
|
SCAR,
|
|
_oracles,
|
|
corpus,
|
|
generate,
|
|
)
|
|
|
|
PLATFORMS = ("discord", "slack", "telegram", "whatsapp")
|
|
|
|
|
|
def test_corpus_ids_unique_and_categorized():
|
|
rows = corpus()
|
|
ids = [r["id"] for r in rows]
|
|
assert len(ids) == len(set(ids)), "vector ids must be unique (runner keys on them)"
|
|
assert {r["category"] for r in rows} == {"grid", "scar", "adversarial"}
|
|
assert len(GRID) >= 20 and len(SCAR) >= 10 and len(ADVERSARIAL) >= 8
|
|
|
|
|
|
def test_oracles_importable_and_self_free():
|
|
"""The generator invokes format_message unbound (self=None) — verify every
|
|
oracle actually renders under that calling convention. A refactor that
|
|
adds instance state to a renderer must update the generator too."""
|
|
oracles = _oracles()
|
|
assert set(oracles) == set(PLATFORMS)
|
|
for platform, render in oracles.items():
|
|
out = render("**bold** and `code`")
|
|
assert isinstance(out, str) and out, platform
|
|
|
|
|
|
|
|
|
|
def test_vector_files_shape(tmp_path):
|
|
generate(tmp_path)
|
|
for platform in PLATFORMS:
|
|
doc = json.loads((tmp_path / f"{platform}.json").read_text(encoding="utf-8"))
|
|
assert doc["platform"] == platform
|
|
assert doc["oracle"]["repo"] == "NousResearch/hermes-agent"
|
|
assert re.match(r"^[0-9a-f]{40}$|^unknown$", doc["oracle"]["commit"])
|
|
assert doc["oracle"]["generator_version"] >= 1
|
|
ids = set()
|
|
for v in doc["vectors"]:
|
|
assert v["expect"] in ("parity", "semantic", "divergent"), v["id"]
|
|
assert isinstance(v["native_output"], str)
|
|
if v["expect"] == "divergent":
|
|
assert v.get("note"), f"{platform}/{v['id']}: divergent needs a note"
|
|
ids.add(v["id"])
|
|
assert len(ids) == len(doc["vectors"])
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def test_committed_vectors_match_regeneration(tmp_path):
|
|
"""The committed tests/conformance/vectors/ must reproduce from the
|
|
current oracle — the same lockstep discipline as openapi.json (drift
|
|
means someone changed a renderer without regenerating)."""
|
|
committed_dir = REPO_ROOT / "tests" / "conformance" / "vectors"
|
|
if not committed_dir.exists():
|
|
pytest.skip("no committed vectors in this checkout")
|
|
generate(tmp_path)
|
|
for platform in PLATFORMS:
|
|
fresh = json.loads((tmp_path / f"{platform}.json").read_text(encoding="utf-8"))
|
|
committed = json.loads((committed_dir / f"{platform}.json").read_text(encoding="utf-8"))
|
|
# Compare everything except the commit stamp (committed file may be
|
|
# one commit behind HEAD in a dirty working tree).
|
|
fresh["oracle"].pop("commit")
|
|
committed["oracle"].pop("commit")
|
|
assert fresh == committed, (
|
|
f"{platform} vectors drifted — run "
|
|
"`python scripts/generate_conformance_vectors.py` and commit"
|
|
)
|