sync: add glm and webextrator service docs from upstream Docs repo

[Copilot]

Upstream AceDataCloud/Docs updated guides for glm, openai, veo, and webextrator. The glm and webextrator service directories were entirely absent from this monorepo.

New services added

glm/

• README.md — service overview, quick-start curl snippet, API table
• docs/glm_chat_completions_api_integration_guide.md — covers all five models (glm-5.1 → glm-3-turbo), basic usage, streaming (SSE), multi-turn, system prompt, function calling, model selection table, error handling

webextrator/

• README.md — service overview, quick-start, API table
• docs/webextrator_render_api_integration_guide.md — full parameter reference, sync/async (callback) modes, error codes
• docs/webextrator_extract_api_integration_guide.md — extraction types (markdown/article/text/links/structured), LLM post-processing, custom instruction
• docs/webextrator_tasks_api_integration_guide.md — single/batch task query, field reference; endpoint is free/unmetered

No-change services

openai/ and veo/ docs were verified to already match upstream content.

[acedatacloud-dev]

Auto-closed during PR backlog cleanup: superseded sync (+667/-0) — content covered by next dispatch against current Docs. Root cause was a broken auto-merge in sync-from-docs.yml (now fixed in PR #143). The fix lands a pre-cleanup step + reliable completion signal so future dispatches will not leak PRs. If this PR contained content not yet in main, the next docs-updated dispatch will re-create it via the working pipeline.