slack-reply-drafting

Draft Slack replies from available context. Use when the user wants help finding messages that likely need a response and preparing reply drafts.

npx skills add https://github.com/openai/plugins --skill slack-reply-drafting

Slack Reply Drafting

Use this skill to identify messages that likely need a reply and produce Slack-ready draft responses from the available context.

Related Skills

WorkflowSkill
Refine, draft, or send the final Slack text../slack-outgoing-message/SKILL.md

Start Here

  • If the user provided channels, threads, DMs, people, or topics, use that scope instead of the default search.
  • If no source scope was provided, default to searching:
    • unanswered direct conversations
    • direct mentions
    • threads with prior user participation and newer replies
    • threads with prior user mention and newer replies
  • For time-specific requests, resolve the user's timezone with slack_read_user_profile.

Support Boundaries

  • This skill is for draft-first reply workflows.
  • If the user explicitly asks to send a reply now rather than prepare a draft, gather the needed Slack context here if useful, then switch to ../slack-outgoing-message/SKILL.md and send directly.
  • Do not invent facts, commitments, approvals, or decisions. If the context is not enough to answer confidently, draft a clarifying reply instead of guessing.

Workflow

  1. Resolve the current user with slack_read_user_profile so you have the user's user_id and can resolve the time window if necessary.
  2. Resolve the time window if the user supplied one.
  3. If the user provided an explicit scope, use the cheapest matching path:
    • specific thread: slack_read_thread
    • named channel: slack_search_channels, then slack_read_channel
    • named person or DM: slack_search_users, then slack_search_public_and_private
    • bounded keyword search: slack_search_public_and_private
  4. If no scope was provided, search these default categories:
    • unanswered direct conversations: slack_search_public_and_private across im,mpim to generate candidate conversations, then slack_read_channel for each plausible candidate before deciding whether it needs a reply; do not decide from the search snippet alone
    • direct mentions: slack_search_public_and_private with query set to <@USER_ID>
    • threads with prior user participation: slack_search_public_and_private with query set to from:<@USER_ID> is:thread, then slack_read_thread for newer replies
    • threads with prior user mention: slack_search_public_and_private with query set to <@USER_ID> is:thread, then slack_read_thread for newer replies after the mention
  5. Keep only candidates where the latest unresolved ask is from someone else, or where newer replies appeared after the user's last substantive reply or mention. Do not count emoji-only, acknowledgement-only, or other non-answer chatter from the user as a reply.
  6. Expand only the threads or surrounding messages needed to answer accurately. Answer the question first, then add clarification or next steps when the context supports it.
  7. If the context is incomplete, write the smallest useful clarifying reply instead of pretending the answer is known.
  8. Finish according to the user's explicit intent:
    • draft/review-first flow: create the draft with slack_send_message_draft in the source channel or DM
    • explicit send-now flow: switch to ../slack-outgoing-message/SKILL.md and send directly Include thread_ts only for thread replies; otherwise omit the parameter entirely. If Slack returns draft_already_exists, stop and tell the user you cannot overwrite the existing attached draft via API.

Drafting Rules

Formatting

  • For a concise Slack or chat summary, you MUST use exactly this structure unless the user explicitly requests a different format.
  • If you use ../slack-outgoing-message/SKILL.md to draft or send the final message, this output contract remains binding. The downstream skill does not relax or rename these sections.

Format multiple drafts as:

**Reply Drafts — <scope>**

**<channel / DM / thread info>**
Draft: <link to draft>

**<channel / DM / thread info>**
Draft: <link to draft>
  • Keep each item minimal: a short header plus the draft link.
  • The header should identify the channel, DM, or thread.
  • If the user asked for a single reply, return just that item.
  • If no unreplied messages are found, say so directly and explain the scope checked.

More skills from openai

user-context
openai
Load or manage the Data Analytics plugin's durable source-routing preferences, onboarding logic, setup progress, and semantic-layer registry.
official
notion-research-documentation
openai
Research Notion content and synthesize into structured briefs, reports, or comparisons with citations. Search and fetch Notion pages using targeted queries, then organize findings by theme with inline source citations and a references section Choose from four output formats (quick brief, research summary, comparison, comprehensive report) based on scope and user goal Create and update Notion pages using built-in templates; link sources directly and track changes as new information arrives...
official
rcsb-pdb-skill
openai
Submit compact RCSB PDB requests for core metadata, Search API queries, and FASTA downloads. Use when a user wants concise RCSB summaries; save raw JSON or…
official
pdf
openai
PDF reading, creation, and validation with visual rendering and programmatic generation. Render PDF pages to PNG for visual inspection of layout, spacing, and typography before delivery using Poppler ( pdftoppm ) Generate PDFs programmatically with reportlab for reliable formatting; extract text and metadata with pdfplumber or pypdf Enforce quality standards: no clipped text, overlapping elements, broken tables, or rendering artifacts; ASCII hyphens only, human-readable citations Use...
official
test-coverage-improver
openai
Improve test coverage in the OpenAI Agents JS monorepo: run `pnpm test:coverage`, inspect coverage artifacts, identify low-coverage files and branches, propose…
official
playwright
openai
Terminal-driven browser automation with element snapshots and interactive UI workflows. Operates via playwright-cli wrapper script (requires npx ); supports headless and headed modes for visual debugging Core workflow: open page, snapshot for stable element references, interact using refs, re-snapshot after navigation or DOM changes Includes form filling, clicking, typing, multi-tab management, screenshot/PDF capture, and trace recording for flow debugging Element refs (e.g., e3 , e15 )...
official
ukb-topmed-phewas-skill
openai
Fetch compact UKB-TOPMed PheWAS summaries for single variants by accepting rsID, GRCh37, or GRCh38 input and resolving to the required GRCh38 query. Use when a…
official
code-review-context
openai
Model visible context
official