Description

Description

On NemoClaw v0.0.38 (OpenClaw v2026.4.24), the Telegram provider's
initialization API calls (deleteWebhook, setMyCommands, deleteMyCommands)
all fail with "Network request for 'X' failed!" immediately after provider
startup. The errors repeat through all retry attempts and never recover.

Core message send/receive functionality works — polling eventually connects
and the bot can receive and reply to user messages. But the initialization
failures mean:
1. Old webhooks are not cleaned up (risk of polling/webhook conflict)
2. Bot slash command menu is not registered
3. Gateway logs are flooded with ERROR-level messages, misleading
   operators into thinking Telegram is broken

Timeline from logs:
  10:36:14 — gateway ready
  10:36:38 — [telegram] starting provider (24s delay after gateway ready)
  10:36:39 — deleteMyCommands FAILED (less than 1s after provider start)
  10:36:39 → 10:37:35 — all retries fail
  10:42:40 — polling receives first user message (6 min after start)

Manual verification from inside the same sandbox:
  node https.request to api.telegram.org/deleteWebhook → OK
  node fetch to api.telegram.org/deleteWebhook → OK
Both succeed, proving the proxy/policy is correct. The openclaw Telegram
plugin uses a different HTTP code path that fails during startup but
works later for message send/receive.

v0.0.37 did not have this issue on the same Brev instance.

Device:        Brev n2d-standard-4
OS:            Ubuntu 22.04.5 LTS
Architecture:  x86_64
Node.js:       v22.22.2
NemoClaw:      v0.0.38
OpenClaw:      v2026.4.24
OpenShell CLI: 0.0.36
Docker:        28.2.2

1. Fresh onboard with Telegram token on Brev:
   TELEGRAM_BOT_TOKEN=... TELEGRAM_USER_ID=... nemoclaw onboard --fresh

2. Check gateway logs immediately after sandbox is ready:
   nemoclaw my-assistant connect
   cat /tmp/gateway.log | grep telegram

3. Observe repeated errors:
   [telegram] deleteMyCommands failed: Network request for 'deleteMyCommands' failed!
   [telegram] deleteWebhook failed: Network request for 'deleteWebhook' failed!
   [telegram] setMyCommands failed: Network request for 'setMyCommands' failed!

4. Wait 5+ minutes, send a message to the bot in Telegram

5. Bot eventually responds (polling works), but initialization
   errors never recover

Telegram provider initialization (deleteWebhook, setMyCommands) should
succeed, or the provider should retry these operations after the proxy
becomes available, rather than exhausting retries during the startup
window when the proxy is not yet ready.

All initialization API calls fail with "Network request failed" and
never retry after the initial retry window. Logs are flooded with
ERROR-level messages. Core messaging works after a delay.

Gateway log (repeating pattern):
[telegram] [default] starting provider
[telegram] deleteMyCommands failed: Network request for 'deleteMyCommands' failed!
[telegram] deleteWebhook failed: Network request for 'deleteWebhook' failed!
[telegram] webhook cleanup failed: Network request for 'deleteWebhook' failed!; retrying in 14.3s.
[telegram] setMyCommands failed: Network request for 'setMyCommands' failed!
[telegram] command sync failed: HttpError: Network request for 'setMyCommands' failed!
(repeats with increasing backoff until retries exhausted)

See Actual Result above. Full timeline:
10:36:07 gateway starting (websocket 1006 errors during startup)
10:36:14 gateway ready (7.2s)
10:36:38 [telegram] starting provider (24s after gateway ready)
10:36:39 first Network request failed
10:37:35 last retry exhausted
10:42:40 first user message received via polling (works)
10:44:47 pairing approved, bot replies to messages (works)

Regression: v0.0.37 on same Brev instance had no Telegram init errors.

Bug Details

[NVB#6164258]