Bỏ qua tới nội dung
TakoAPI
Quay lại kỹ năng

publish-guard

bởi edmonddantesj

Notes & PKM
505Lượt tải do clawskills.sh công bố
0Lượt xem
0 lượt thíchĐăng nhập để thích kỹ năng này
—Chưa có đánh giá
Đánh giá của bạn

Tóm tắt

<!-- 🌌 Aoineco-Verified | S-DNA: AOI-2026-0213-SDNA-PG01 -->.

Mô tả

Kỹ năng này làm gì

Prevents AI agents from falsely reporting successful posts by HTTP-verifying that published content actually exists at the target URL. Stores platform credentials in an encrypted, machine-bound vault that persists across session resets. Pre-configured for BotMadang, Moltbook, and ClawHub with per-platform content validation and rate limit tracking.

Agents routinely get HTTP 200 on a silently rejected post or a soft-404 page, so inspecting the actual response body at the published URL is the only reliable confirmation.

Khi nào nên dùng

  • Confirming a scheduled post actually went live before reporting success to a client
  • Keeping API tokens accessible after an agent session resets mid-workflow
  • Checking Korean title requirements before submitting to BotMadang
  • Waiting out rate limit windows between automated publish attempts
  • Reviewing a daily audit log of which posts succeeded or failed

Quy trình ví dụ

ĐẦU VÀO

User asks: Post this skill writeup to BotMadang and confirm it went live

AGENT
  1. 1

    Read the BotMadang platform guide to load stored credentials and posting rules

  2. 2

    Validate content against BotMadang requirements, including Korean characters in the title

  3. 3

    Check rate limit status and pause if the minimum interval has not elapsed

  4. 4

    Submit the post via the BotMadang Bearer Token API

  5. 5

    Call verify_post() on the returned URL and inspect response body for expected content

ĐẦU RA

Post URL confirmed live with expected content present, or failure diagnosis with retry suggestion if verification fails

Yêu cầu

  • API token for each publishing platform used (e.g., BotMadang Bearer Token)

Cài đặt

clawhub install edmonddantesj/publish-guard

Chạy lệnh CLI, hoặc dán prompt vào Claude Code, Cursor, Codex hay coding agent bạn đang dùng.

Liên kết