Telegram 媒体拉取走系统代理：补齐 HTTPS_PROXY/HTTP_PROXY/ALL_PROXY 回退

问题/场景：企业/WSL 环境只配环境变量代理时，Telegram 媒体下载 fetch failed。前置条件：未显式设置 channels.telegram.proxy，但已设置代理环境变量。实施步骤：实现 resolveProxyUrlFromEnv（优先级 HTTPS→HTTP→ALL，大小写不敏感）→ monitor 路径回退到 env 代理 → 保留显式配置优先。关键命令：HTTPS_PROXY=http://proxy:3128 openclaw start。验证：媒体下载请求改经代理转发。风险：send/probe/audit 等链路可能未完全覆盖（需验证）。来源：PR #20883。

GITHUBDiscovered 2026-02-19Author arosstale

Prerequisites

Runtime uses Node.js fetch with proxy env vars set at process level.
Telegram account has no explicit per-account proxy override (for fallback test).

Steps

Export proxy env vars before startup (HTTPS_PROXY preferred).
Implement env resolver with precedence: HTTPS_PROXY → HTTP_PROXY → ALL_PROXY (case-insensitive).
In Telegram monitor/media path, use explicit channels.telegram.proxy first; fallback to env resolver only when absent.
Re-run media upload/download scenario and compare with baseline failure logs.

Commands

HTTPS_PROXY=http://corporate-proxy:3128 openclaw start

openclaw gateway status

Verify

MediaFetchError fetch failed disappears and media operations succeed via proxy env setup.

Caveats

Explicit channel proxy must remain highest priority to avoid unexpected route changes.
send/probe/audit paths may still need same fallback for full consistency（需验证）.

Source attribution

This tip is aggregated from community/public sources and preserved with attribution.

Open original source ↗

Visit original post