Skills Plugins MCP Prompt Model 博客 我的中心

release-changelog-discord-message

Write the Discord announcement for a stable Paperclip release from the release changelog. Use when a release issue needs a copy-pasteable dotta-voice Discord post or refreshed discord_announcement document.

DeepseekModel 官方收录技能 质量 优秀 · 90 v1.0.0

获取

https://deepseekmodel.com/api/download.php?id=paperclipai-paperclip-agents-skills-release-changelog-discord-message-skill-md&format=skill
下载 .skill 标准格式,含 system_prompt 与 model_config,导入任意 Agent 框架即可使用
.skill 文件中 system_prompt 字段的实际内容。
name release-changelog-discord-message description Write the Discord announcement for a stable Paperclip release from the release changelog. Use when a release issue needs a copy-pasteable dotta-voice Discord post or refreshed discord_announcement document. Release Discord Announcement Skill Write the Discord release announcement for the stable Paperclip release. This is the companion to .agents/skills/release-changelog/SKILL.md . That skill writes the changelog — during the beta soak it lives at releases/beta/v{beta-version}.md on the release-notes/v{beta-version} branch, and after the stable ships a canonicalization PR renames it to releases/vYYYY.MDD.P.md (see that skill's Channel Process section). This skill turns that file into a single copy-pasteable Discord block, in dotta's voice, and posts it as the discord_announcement document on the release issue. What dotta said This is for discord — try to follow my format. If I have a section where I think about the future, pull from recent issues we're working on etc. The Discord announcement is not the changelog. The changelog is exhaustive; the announcement is opinionated, in-voice, and built around the same handful of shipped highlights plus a real "what's next" + "what's on my mind" pulled from current Paperclip work — not invented. When to use After release-changelog has produced the changelog (beta-keyed on the release-notes/v{beta-version} branch during the soak, or the canonicalized releases/vYYYY.MDD.P.md after the stable ships). When the release issue (the one assigned by the release routine) asks for a Discord announcement, or has a discord_announcement document that needs to be refreshed for a new date/version. Never run this in isolation. The version, date, contributor list, and highlight set MUST match the matching changelog file — if the changelog has been updated, refresh this too. Output A single fenced markdown code block, ready to paste into Discord. Attached as issue document key discord_announcement on the release issue, and pasted verbatim into a comment on that issue so the human can copy it out. When Cases are enabled, also upsert the social child case described below. PUT /api/issues/{releaseIssueId}/documents/discord_announcement { "title" : "Discord announcement" , "format" : "markdown" , "body" : "<the announcement>" , "baseRevisionId" : "<latest if updating>" } If the document already exists, fetch it first and pass the current baseRevisionId . Never overwrite silently — if the version has changed since the document was last written, mention what changed in the issue comment. Format (follow this template) Use Discord emoji shortcodes ( :paperclip: , :lock: , :brain: …) — NOT the Unicode emoji. Discord renders the shortcodes; the changelog file uses prose. :paperclip: :paperclip: :paperclip: CLIPPERS!!! v{VERSION} IS OUT :paperclip: :paperclip: :paperclip: OFFICIAL TWITTER: https://x.com/papercliping - follow it, report any others ## Highlights :emoji: **Feature Name** - one-sentence description in dotta's voice. :emoji: **Feature Name** - … :emoji: **Feature Name** - … ... and a long tail of {flavor of the rest}. Read the [full release notes](<github link>). ## WHATS NEXT (:motorway: Roadmap) * **Theme A** - one-line forward-looking blurb * **Theme B** - … * **Theme C** - … ## What's on my mind * **Topic** - what's bugging dotta / what's queued / open questions * **Topic** - … ## PRESS (optional — only if there is real press) * **Outlet / Person** - what happened ([link](<x.com link>)) ## WHAT I NEED FROM YOU (optional — only if there's a real ask) FOLLOW THE TWITTER: https://x.com/papercliping - that's the only official one TELL ME if you're using Paperclip in your business - I want to meet you ## Community Thank you to everyone who contributed to this release! @username1, @username2, @username3 ## In Summary PAPERCLIP IS THE AI ORCHESTRATOR FOR HUMANS TO ACCOMPLISH 100x MORE WORK Every single person will be managing a team of a dozen, or a hundred, or a thousand agents and Paperclip will be the default tool to manage it all. ITS TIME TO CLIP :paperclip: :paperclip: :paperclip: FULL RELEASE NOTES https://github.com/paperclipai/paperclip/blob/master/releases/v{VERSION}.md ||@everyone|| Notes on the template: The opening and closing :paperclip: :paperclip: :paperclip: bookends are part of the brand — keep them. Name the install channels somewhere in the post: npx paperclipai@latest for the stable, @beta / @nightly / @canary for earlier access, and Docker :latest moving only on stable releases. The FULL RELEASE NOTES link points at releases/v{VERSION}.md on master — that file exists only after the post-stable canonicalization PR merges. Merge it before the announcement is posted. Sections may be UPPERCASE or Title Case — dotta has used both. Pick a style and stay consistent within a single post. Use ||@everyone|| (Discord spoiler-wrapped) at the very end so it pings exactly once when the spoiler is removed by the poster. Language tips These are extracted from how dotta has written the last several announcements. Mimic this register; do not invent a "professional" tone. First person, conversational. "I want to meet companies using Paperclip", "what's on my mind", "if that's you let me know". Not "Paperclip is excited to announce". ALL CAPS for excitement and asks , especially in the opener, the section headers, the "WHAT I NEED FROM YOU" section, and the closing tagline. Do not ALL-CAPS feature descriptions. One emoji shortcode per highlight bullet , picked to evoke the feature ( :lock: for secrets, :brain: for planning, :mag: for search, :cloud: for cloud / sandbox, :jigsaw: for plugins, :rewind: for history/restore, :thread: for threads, etc.). Highlight bullets are one sentence , opinionated, told from the user's perspective — "the cloud-secrets prereq is real now", not "added support for…". Tail line after highlights wraps the rest in a single sentence and links to the full release notes ("… and a long tail of {flavor}. Read the full release notes ."). "WHATS NEXT" is forward-looking themes , not a literal sprint list. 3–5 bullets is the right size. Pull these from active goals, in-flight projects, and recent issues the team is working on — do not invent themes. Highlights follow the changelog skill's delta rule : a feature the previous announcement already introduced appears only for what changed this release, and does not headline twice for follow-through work. "What's on my mind" is dotta's personal/strategic thinking — docs gaps, philosophical positioning ("we're the human control plane for ai labor"), invitations ("if you've ever wanted to write about how you use Paperclip, hit me up"). Pull real tensions from recent issues/comments; do not invent. Press section is optional. Only include it if there is real press in the release window (a tweet, a podcast, a talk, a star milestone). No press → drop the section entirely. "WHAT I NEED FROM YOU" is optional. Use it for a single concrete ask (follow the twitter, intros, beta sign-ups). No real ask → drop it. Community is the same contributors list that's in the changelog file, fenced in a triple-backtick block, comma-separated @username, @username . Exclude bots and the specific folks on the changelog skill's canonical exclusion list — same rules. The "In Summary" mission line evolves slowly. Use the most recent variant unless dotta tells you otherwise. Recent variants: "PAPERCLIP IS THE AI ORCHESTRATOR FOR HUMANS TO ACCOMPLISH 100x MORE WORK" "PAPERCLIP WILL BE THE DEFAULT AGENT-MANAGEMENT TOOL FOR EVERY COMPANY" "Paperclip will be the control plane for AI agents in every company." Closing tagline is always ITS TIME TO CLIP :paperclip: :paperclip: :paperclip: . Keep it. Workflow Read the matching changelog produced by release-changelog — the beta-keyed file during the soak, releases/vYYYY.MDD.P.md once canonicalized. Use the version and contributor list from that file — never re-derive them. Resolve the parent release case with key paperclip-release:vYYYY.MDD.P . If it does not exist and Cases are enabled, create it using the schema in .agents/skills/release-changelog/SKILL.md before creating child cases. Read the release issue thread (the one assigned to you that ran the release routine) — comments + linked issues + recent issues in the company are the source for WHATS NEXT and What's on my mind . Commits already on origin/master after the beta source commit are prime "what's next" material: they are literally the next release's content. Pull real themes, not invented ones. Re-read the three verbatim examples below — they're the canonical voice. Draft the announcement using the template above. PUT it as the discord_announcement document on the release issue (see "Output" above). If updating, send the latest baseRevisionId . Upsert the tweet_storm child case with parentCaseId set to the release case id, then PUT its body document to the announcement body. Post a comment on the release issue that includes the announcement inside a single fenced markdown code block, so dotta can copy-paste it into Discord without opening the document. Tweet Storm Case Schema Use this child case for the Discord/social announcement thread. The key must be stable so retries update the same child case: POST /api/companies/:companyId/cases { "caseType": "tweet_storm", "key": "paperclip-release:vYYYY.MDD.P:tweet-storm", "title": "Paperclip vYYYY.MDD.P tweet storm", "summary": "Social announcement thread for Paperclip vYYYY.MDD.P.", "status": "in_review", "parentCaseId": "<release-case-id>", "fields": { "schema_version": 1, "version": "vYYYY.MDD.P", "channel": "x", "discord_source": true, "post_count": 1, "target_audience": ["operators", "contributors", "agent-company builders"], "links": { "release_notes": "https://github.com/paperclipai/paperclip/blob/master/releases/vYYYY.MDD.P.md", "official_account": "https://x.com/papercliping" }, "review": { "needs_human_copy_paste": true, "approved_by": null } } } Then write the body document: PUT /api/cases/:tweetStormCaseId/documents/body { "title": "Paperclip vYYYY.MDD.P tweet storm body", "format": "markdown", "body": "<announcement body>", "changeSummary": "Draft social announcement" } If updating an existing document, fetch the case and pass the latest baseRevisionId . Do not publish to Discord. This skill only prepares the artifact. Verbatim previous examples Three previous Discord announcements from dotta, included verbatim as the ground-truth examples for voice, structure, and emoji usage. When in doubt, match these. Example 1 — v2026.403.0 CLIPPERS! v2026.403.0 has dropped!! :paperclip: :paperclip: :paperclip: ## Highlights :inbox_tray: **Inbox overhaul** - there is a new "mine" tab that has mail-client like keyboard shortcuts. It's my new default view for managing work :thumbsup: **Feedback and evals** - you can now vote :thumbsup: / :thumbsdown: on your agent's responses. If you choose to share your traces with me, I'll use it to make Paperclip better. In either case you can export locally for your own org's learning :page_with_curl: **Document revisions** - you can now restore old versions of your documents :ping_pong: **Telemetry** - this version has anonymized telemetry that helps me better understand the basic uses of Paperclip (adapters and so on) - if you hate that, just it disable with `DO_NOT_TRACK=1` or `PAPERCLIP_TELEMETRY_DISABLED=1` environment variables :construction_worker: **Execution Workspaces (experimental)** - Paperclip is not a "code review" tool, but I have been finding worktrees are important for certain projects. Enable it in experimental settings :loop: **Routine variables** - sometimes you need to customize a routine and the new variables feature makes that easy PLUS **tons** of improvements aound adapters, bugfixes, qol ## COMMUNITY HUGE THANKS to the contributors with commits in this release: @aronprins, @bittoby, @edimuj, @HenkDz, @kevmok, @mvanhorn, @radiusred, @remdev, @statxc, @vanductai ## WHATS NEXT (ROADMAP) * **Multi-human users** -- you've been asking for it, we have a draft and will have this shortly * **Sandbox execution** - the other half of cloud deployment: run your agents in a sandbox across any provider PLUS: just dealing with the excellent PRs we have sitting in our inbox. **What's also on my mind (coming soonish)** * MAXIMIZER MODE - for when you've got a dream and tokens to burn * Artifacts, work products, and deployments * CEO Chat * Stronger agent defaults ## PRESS I've been doing my part to spread the word about Paperclip * We talked to the incredible [Andrew Warner of Mixergy Fame](https://x.com/dotta/status/2039087507514507407) * We gave a tutorial with the [inimitable Greg Isenberg](https://x.com/dotta/status/2037279902445994345) * We met with the [Seed Club guys](https://x.com/dotta/status/2039020365926576377) * We crossed [40k stars (46k now!)](https://x.com/dotta/status/2038638188227387613) * ... and a couple others that will be released in a few days ## SUCCESS STORIES * [Nevo made $76k in march](https://x.com/dotta/status/2039406772859920758) after using Paperclip to automate his marketing * [Lewis Jackson](https://x.com/WhatSayLew/status/2039810227394978158) said 34 agents were already operating his trading firm through Paperclip and called it his "holy s***" AI moment. * [Neal Kotak](https://x.com/nkotak1/status/2039582439459209638) said Paperclip already runs most of Roominary for him and praised how strong the product is. * [Sam Woods](https://x.com/samwoods/status/2039039305960587755) said he knows several people who moved from OpenClaw to Paperclip, often with Hermes in the stack, and that they love it. * [Josh Galt](https://x.com/JoshGalt/status/2039386307219095557) called Paperclip the coolest agent tooling he has used and said it is finally something that just works. ## IN SUMMARY I know there are still some rough edges, but Paperclip will be *the* control plane for AI agents in **every** company. and I think we're moving at a pretty good clip :paperclip: :paperclip: :paperclip: FULL RELEASE NOTES HERE https://github.com/paperclipai/paperclip/releases/tag/v2026.403.0 ||@everyone|| Example 2 — v2026.416.0 :paperclip: :paperclip: :paperclip: CLIPPERS!!! v2026.416.0 IS OUT :paperclip: :paperclip: :paperclip: ## Highlights This release has *tons* of quality of life improvements around speed, performance, and workflow. You should notice that comment threads feel faster and your agents stay on task longer :thread: Issue chat threads now are a conversation more than comments :police_officer: Execution policies like **Reviewer** and **Approver** are now first-class in the harness (e.g. enforce that QA *must* review a task) :no_smoking: Blocker dependencies - first-class "wake on blocker resolved" which means now you can have "task graphs" that depend on one another and it's enforced by Paperclip :woman_feeding_baby: Parent-child tasks - better support for sub-tasks all around, which makes it much easier to organize your work And then a million fixes around ux, details, keyboard shortcuts, bug fixes, security fixes, etc. Really you should read the [full release notes here](https://github.com/paperclipai/paperclip/releases/tag/v2026.416.0) ## COMMUNITY INCREDIBLE INCREDIBLE WORK BY folks with commits and reports in this release: @AllenHyang, @antonio-mello-ai, @aronprins, @chrisschwer, @cleanunicorn, @DanielSousa, @davison, @ergonaworks, @HearthCore, @HenkDz, @KhairulA, @kimnamu, @Lempkey, @marysomething99-prog, @mvanhorn, @officialasishkumar, @plind-dm, @shoaib050326, @sparkeros, @wbelt, @offset, @sagilayani, @mattdonnelly10, @peaktwilight, @YuvalElbar6 ## WHATS NEXT (:motorway: Roadmap) * **Multi-human users** - in the last stages of testing, Paperclip is better with teams * **Memory Infrastructure** - your agents will remember everything about yoru business * **Sandbox execution** - run your agents anywhere ## What's on my mind * I want to meet with companies who are using Paperclip in their business - if that's you let me know * We need more Paperclip tutorials, defaults, and education - thanks to @aronprins for his work in this area already! * We still need to get better at reviewing your PRs and we're improving our process every day * "Zero-human company" language has to go - we're the human control plane for ai labor * We're adding better support for *knowledge (wikis & files)*, *artifacts*, and *work product* in Paperclip soon. ## PRESS
Agent 识别该技能的关键词,点击任意一个即可复制。

该技能未提供触发词。

下载的 .skill 包内含以下字段。
字段 说明
format格式标识(skill/v1)
skill_id技能唯一 ID
name技能名称
version版本号
description技能描述
category所属分类(数组)
trigger_words触发词列表
tags标签列表
source来源标识
source_url来源链接(本页地址)
exported_at导出时间(每次下载生成)
system_prompt系统提示词正文
model_config模型参数:provider / model / temperature / max_tokens / top_p
examples示例
install_guide各平台导入说明(Coze / Dify / Claude / 自定义框架)
同一份技能可按不同平台格式导出。
.skill 标准格式,含 system_prompt 与 model_config,导入任意 Agent 框架即可使用 下载
.skillpro 增强格式,额外含脚本 / 工具 / 依赖 / 钩子占位 下载
.json 纯 JSON 导出,只含 system_prompt 与模型参数 下载
Coze 带 frontmatter 的 Markdown,Coze 平台导入用 下载
Dify Dify DSL,创建应用后直接导入 下载

每日精选 Skill 推荐,免费送到你邮箱

输入邮箱,每天接收一个精选 AI Agent 技能推荐。完全免费,持续更新。

提交后我们会发送一封确认邮件,点击邮件里的链接才会开始收信。

完全免费,取消任意时间。我们不会发送垃圾邮件。