Inbox to Action Closer — Inbox to Action — 收件箱处理

Purpose

Process raw pending-work data from multiple sources, deduplicate and score the items, and present one unified action board. This skill does not fetch data from external APIs. It expects raw source data to be supplied by the caller, by other OpenClaw skills, or by tools that have already retrieved it. All output is draft-only until the user explicitly confirms a write action.

Data Acquisition

This skill does not include API connectors or manage credentials. To use it:

• Supply raw JSON data 从 每个 source (Slack messages, GitHub PRs, 日历 events, etc.)
• 使用 existing OpenClaw tools 或 installed skills 已经 connect 到 these services
• 或 pipe 输出 从 CLI tools (gh, himalaya, slack-cli, etc.) 进入 normalisation pipeline

The skill handles everything after data retrieval: normalisation, deduplication, scoring, and rendering.

Supported Source Formats

• Slack — 消息 和 thread JSON (sender, channel, 时间戳, participants, permalink)
• GitHub — PR, issue, 和 review 请求 JSON (assignee, title, state, URL)
• 日历 — 事件 JSON (summary, 开始/end 乘以, attendees, location)
• Notion — page 或 task JSON (title, status, assignee, due 日期, URL)
• Trello — card JSON (name, 列表, members, due 日期, URL)
• Email — 消息 JSON (从, subject, 日期, flags, thread references)

Execution Steps

• 接收 raw source data 从 caller 或 upstream tools.
• 对于 每个 source, call corresponding adapter 通过 normalize (src/normalize.ts) 到 convert raw items 进入 normalised action-item schema defined 在...中 src/types.ts. 如果 source missing, skip cleanly.
• Pass 所有 normalised items 通过 dedupe (src/dedupe.ts) 到 merge cross-source duplicates 使用 conservative confidence-based matching.
• Score every item 使用 score (src/score.ts) 到 compute transparent urgency rankings.
• Generate final action board 使用 render (src/render.ts) 在...中 both markdown 和 structured JSON formats.
• orchestration entrypoint 索引 (src/索引.ts), 哪个 coordinates steps 1-5.

Safety Rules

• 所有 输出 必须 draft-仅 由 默认. NEVER auto-发送 messages, post comments, 关闭 issues, 或 perform 任何 写入 action 没有 explicit 用户 confirmation.
• NEVER auto-post 到 任何 source system. Generated 回复 drafts 和 suggested actions proposals, 不 executions.
• NEVER perform destructive actions such 作为 deleting items, archiving threads, 或 dismissing notifications.
• ALWAYS ask 用户 对于 explicit confirmation 之前 executing 任何 写入 action, 包括 sending replies, posting comments, updating task statuses, 或 creating 新的 items. confirmation gate mandatory 和 separate 从 draft-仅 默认.
• 如果 source 不可用, misconfigured, 或 returns 错误, skip cleanly 和 continue 处理中 remaining sources. NEVER 失败 entire run 因为 one source unreachable.
• 必须 不 fabricate 或 hallucinate action items. 仅 surface items exist 在...中 source data.
• 必须 preserve original source URL 对于 every action item 所以 用户 可以 验证 和 act 在...中 context.