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.
PTB's HTTPXRequest builds its httpx.AsyncClient with
`limits = httpx.Limits(max_connections=connection_pool_size)` and no
keepalive tuning, so httpx's default keepalive_expiry=5.0 applies. Behind
an HTTP proxy (Cloudflare Warp etc.) a peer-initiated FIN can sit in
CLOSE_WAIT longer than that, leaking fds in the general request pool
(_request[1], which routes bot.send_message/set_my_commands) — the pool
_drain_polling_connections never resets. Telegram was the lone holdout
adapter not using the shared #18451 CLOSE_WAIT helper.
Wire gateway.platforms._http_client_limits.platform_httpx_limits() into
the httpx client across ALL THREE request-construction branches —
fallback-transport, proxy, and plain — via httpx_kwargs["limits"], which
PTB spreads last into its client kwargs so our tuned limits win. PTB's
connection_pool_size (max_connections) is preserved; only keepalive
behaviour is tightened (max_keepalive_connections + keepalive_expiry<5.0).
The fix is macOS-import-safe: no Linux-only socket TCP_KEEPIDLE/INTVL/CNT
constants at module scope (unlike the broken candidate which crashed on
import on the reporter's OS), and it patches the actual proxy path the
repro hits rather than TelegramFallbackTransport, which the proxy repro
never instantiates.
Adds a mutation-survivable behavior-contract test asserting every
HTTPXRequest built by connect() receives httpx_kwargs["limits"] with
keepalive_expiry < httpx's 5.0 default, across both the proxy and plain
branches. Reverting the limits wiring fails the test.
Co-authored-by: indigokarasu <mx.indigo.karasu@gmail.com>
Teknium
kshitijk4poor