Token导航 LogoToken导航

Agent Skill · 开发

nostr-event-builder(Nostr 事件生成器)

nostr-event-builder 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

nostr-event-builder 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在开发类的 21,746 个 Skill 里排第 19,065。

总安装 294周安装 12GitHub Stars 4下载 94核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:nostr-event-builder(Nostr 事件生成器)
来源仓库:https://github.com/accolver/skill-maker
仓库路径:skills/nostr-event-builder
安装命令:
npx skills add https://github.com/accolver/skill-maker --skill nostr-event-builder
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/accolver/skill-maker --skill nostr-event-builder

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息。适合围绕仓库状态、代码变更或协作事项进行整理,提升开发效率。可结合来源仓库和原始 README 核验具体用法,确保适用性。

安装方式:通过 npx 从 GitHub 仓库添加,支持 Codex、Claude、Cursor、Gemini CLI。注意:安装前建议确认权限范围和维护状态,避免触发不必要的联网或文件操作。

来源仓库内路径:skills/nostr-event-builder

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/accolver/skill-maker --skill nostr-event-build…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

nostr-event-builder 怎么安装?

在终端执行 npx skills add https://github.com/accolver/skill-maker --skill nostr-event-builder。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

nostr-event-builder 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

nostr-event-builder 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

nostr-event-builder 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

nostr-event-builder 用的人多吗?

公开数据是累计安装 294 次,近一周安装 12 次,下载 94 次,GitHub 4 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

nostr-event-builder 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理开发类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
nostr-event-builder

Nostr 事件生成器 · 本页

2944只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Nostr Event Builder

Overview

Construct correct Nostr event structures from natural language descriptions. This skill handles the non-obvious parts: choosing the right kind, building proper tag arrays with correct markers, enforcing NIP-10 vs NIP-22 threading rules, and producing valid event JSON ready for signing.

When to Use

  • The task is to construct or validate a concrete Nostr event JSON object and its tag array.
  • The user describes what they want to publish and needs the exact event shape, kind, tags, and content rules.
  • The problem is correctness of event structure for replies, comments, metadata, or other publishable events.
  • The output should be an event payload, not protocol-planning advice.

Do NOT use when:

  • The task is deciding which NIPs apply at a high level.
  • The work is subscription filter design or relay query construction.
  • The problem is relay implementation rather than event payload construction.

Response format

Always structure the final response with these top-level sections, in this order:

  1. Summary — state the task, scope, and main conclusion in 1-3 sentences.
  2. Decision / Approach — state the key classification, assumptions, or chosen path.
  3. Artifacts — provide the primary deliverable(s) for this skill. Use clear subheadings for multiple files, commands, JSON payloads, queries, or documents.
  4. Validation — state checks performed, important risks, caveats, or unresolved questions.
  5. Next steps — list concrete follow-up actions, or write None if nothing remains.

Rules:

  • Do not omit a section; write None when a section does not apply.
  • If files are produced, list each file path under Artifacts before its contents.
  • If commands, JSON, SQL, YAML, or code are produced, put each artifact in fenced code blocks with the correct language tag when possible.
  • Keep section names exactly as written above so output stays predictable across skills.

Workflow

1. Identify the Event Kind

Ask: "What is the developer trying to publish?"

IntentKindCategoryKey NIP
Post a short text note1RegularNIP-10
Reply to a kind:1 note1RegularNIP-10
Comment on non-kind:1 content1111RegularNIP-22
Set/update user profile0ReplaceableNIP-01
Update follow list3ReplaceableNIP-02
Delete events5RegularNIP-09
Repost a note6RegularNIP-18
React to an event7RegularNIP-25
Publish a long-form article30023AddressableNIP-23

See references/event-kinds.md for full kind-to-structure mapping.

Critical routing rule:

Is the target a kind:1 note?
  YES → Use kind:1 reply with NIP-10 e-tag markers
  NO  → Use kind:1111 comment with NIP-22 uppercase/lowercase tags

2. Build the Tag Array

Tags are the hardest part. Follow the tag guide for your kind:

For kind:1 replies (NIP-10):

Direct reply to root (no intermediate replies):

{
  "kind": 1,
  "tags": [
    ["e", "<root-event-id>", "<relay-url>", "root", "<root-author-pubkey>"],
    ["p", "<root-author-pubkey>"]
  ],
  "content": "Your reply text"
}

Reply to a reply in a thread:

{
  "kind": 1,
  "tags": [
    ["e", "<root-event-id>", "<relay-url>", "root", "<root-author-pubkey>"],
    [
      "e",
      "<parent-event-id>",
      "<relay-url>",
      "reply",
      "<parent-author-pubkey>"
    ],
    ["p", "<root-author-pubkey>"],
    ["p", "<parent-author-pubkey>"]
  ],
  "content": "Your reply text"
}

For kind:1111 comments (NIP-22):

Top-level comment on a regular event:

{
  "kind": 1111,
  "tags": [
    ["E", "<root-event-id>", "<relay-url>", "<root-author-pubkey>"],
    ["K", "<root-event-kind>"],
    ["P", "<root-author-pubkey>", "<relay-url>"],
    ["e", "<root-event-id>", "<relay-url>", "<root-author-pubkey>"],
    ["k", "<root-event-kind>"],
    ["p", "<root-author-pubkey>", "<relay-url>"]
  ],
  "content": "Your comment text"
}

Top-level comment on an addressable event (kind 30000-39999):

{
  "kind": 1111,
  "tags": [
    ["A", "<kind>:<pubkey>:<d-tag>", "<relay-url>"],
    ["K", "<root-event-kind>"],
    ["P", "<root-author-pubkey>", "<relay-url>"],
    ["a", "<kind>:<pubkey>:<d-tag>", "<relay-url>"],
    ["e", "<event-id>", "<relay-url>"],
    ["k", "<root-event-kind>"],
    ["p", "<root-author-pubkey>", "<relay-url>"]
  ],
  "content": "Your comment text"
}

Comment on a URL or external identifier:

{
  "kind": 1111,
  "tags": [
    ["I", "<url-or-identifier>"],
    ["K", "<identifier-type>"],
    ["i", "<url-or-identifier>"],
    ["k", "<identifier-type>"]
  ],
  "content": "Your comment text"
}

Reply to an existing comment:

{
  "kind": 1111,
  "tags": [
    ["E", "<original-root-event-id>", "<relay-url>", "<root-author-pubkey>"],
    ["K", "<original-root-kind>"],
    ["P", "<root-author-pubkey>"],
    [
      "e",
      "<parent-comment-id>",
      "<relay-url>",
      "<parent-comment-author-pubkey>"
    ],
    ["k", "1111"],
    ["p", "<parent-comment-author-pubkey>"]
  ],
  "content": "Your reply to the comment"
}

See references/tag-guide.md for complete tag semantics.

3. Set the Content Field

Content format depends on the kind:

KindContent Format
0Stringified JSON: {"name":"...","about":"...","picture":"..."}
1Plaintext (no markdown, no HTML)
5Optional deletion reason text
6Stringified JSON of the reposted event
7+ (like), - (dislike), or emoji
1111Plaintext comment
30023Markdown-formatted article body

4. Construct the Complete Event

Assemble the unsigned event object:

{
  "pubkey": "<32-bytes-lowercase-hex-public-key>",
  "created_at": "<unix-timestamp-seconds>",
  "kind": "<integer>",
  "tags": [["..."]],
  "content": "<string>"
}

The id is computed as SHA-256 of the serialized form:

[0, "<pubkey>", <created_at>, <kind>, <tags>, "<content>"]

Serialization rules:

  • UTF-8 encoding, no whitespace/formatting
  • Escape in content: \n, \", \\, \r, \t, \b, \f
  • All other characters verbatim

The sig is a Schnorr signature (secp256k1) of the id.

5. Validate Before Signing

Checklist before the event is ready:

  • kind is correct for the intent
  • All required tags present for this kind
  • e tags have correct markers (root/reply for kind:1)
  • p tags include ALL participants in the thread
  • NIP-10 kind:1 replies only target other kind:1 events
  • NIP-22 kind:1111 comments do NOT target kind:1 events
  • K and k tags present for kind:1111 comments
  • Uppercase tags (E/A/I/K/P) point to root scope in kind:1111
  • Lowercase tags (e/a/i/k/p) point to parent item in kind:1111
  • content format matches the kind's requirements
  • created_at is a Unix timestamp in seconds (not milliseconds)
  • All hex values are 32-byte lowercase

Common Mistakes

MistakeWhy It BreaksFix
Using kind:1 to reply to a kind:30023 articleNIP-10 kind:1 replies MUST only reply to other kind:1 eventsUse kind:1111 (NIP-22 comment) for non-kind:1 targets
Using kind:1111 to reply to a kind:1 noteNIP-22 comments MUST NOT reply to kind:1Use kind:1 with NIP-10 e-tag markers
Missing root marker on e tags in kind:1Clients can't reconstruct the thread treeAlways use marked e tags: ["e", "<id>", "<relay>", "root"]
Only one e tag with reply marker (no root)Direct replies to root need root marker, not replySingle e tag = use root marker only
Missing p tags for thread participantsUsers don't get notified of repliesInclude p tags for ALL pubkeys in the thread
Lowercase e/k/p tags for root scope in kind:1111Root scope MUST use uppercase E/K/P tagsUppercase = root scope, lowercase = parent item
Missing K or k tags in kind:1111Both are REQUIRED by NIP-22Always include ["K", "<root-kind>"] and ["k", "<parent-kind>"]
created_at in millisecondsNostr uses seconds, not millisecondsUse Math.floor(Date.now() / 1000)
Content as object instead of string for kind:0Content must be stringified JSONUse JSON.stringify({name: "...",...})
Missing d tag on addressable events (30000-39999)Relay can't address the event properlyAlways include ["d", "<identifier>"]
Positional e tags without markersDeprecated; creates ambiguity in thread reconstructionAlways use marked e tags with root/reply

Kind Category Quick Reference

RangeCategoryBehavior
1000-9999, 4-44, 1, 2RegularStored by relays, all kept
10000-19999, 0, 3ReplaceableLatest per pubkey+kind kept
20000-29999EphemeralNot stored by relays
30000-39999AddressableLatest per pubkey+kind+d-tag kept

Example: Complete Event Construction

User says: "I want to reply to my friend's note in an existing thread"

Step 1 — Identify kind: Replying to a kind:1 note → use kind:1 (NIP-10)

Step 2 — Determine thread position: This is a reply to a reply (not the root), so we need both root and reply e tags.

Step 3 — Build the event:

{
  "pubkey": "a1b2c3d4e5f6a7b8c9d0e1f2a3b4c5d6e7f8a9b0c1d2e3f4a5b6c7d8e9f0a1b2",
  "created_at": 1709827200,
  "kind": 1,
  "tags": [
    ["e", "aaa111...", "wss://relay.example.com", "root", "f7234bd4..."],
    ["e", "bbb222...", "wss://relay.example.com", "reply", "93ef2eba..."],
    ["p", "f7234bd4...", "wss://relay.example.com"],
    ["p", "93ef2eba...", "wss://relay.example.com"]
  ],
  "content": "Great point! I totally agree with this take."
}

Step 4 — Validate: Root e tag has root marker ✓, reply e tag has reply marker ✓, p tags include both the root author and the parent author ✓, content is plaintext ✓.

平台分布

Codex

37.47%
按下载量换算35

Claude

27.65%
按下载量换算26

Cursor

20.8%
按下载量换算20

Gemini CLI

10.28%
按下载量换算10

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills