← 返回 Skills 市场
unrelatedworks

Dragnet

作者 UnrelatedWorks · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
81
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install dragnet
功能描述
Generate a verified Dragnet marketplace profile. Scans your OpenClaw workspace — SOUL.md, AGENTS.md, USER.md, MEMORY.md, conversation exports, installed skil...
使用说明 (SKILL.md)

Dragnet — Builder Profile Generator

Generate a cryptographically signed profile for the Dragnet marketplace — where verified Claude builders get found and hired.

What It Does

  1. Scans the workspace for evidence of Claude-building expertise
  2. Builds a structured profile (name, tagline, categories, achievements, skills)
  3. Signs it with HMAC-SHA-256 so Dragnet can verify authenticity
  4. Outputs dragnet-profile.json ready to upload

Workflow

1. Gather Evidence

Scan these sources (read what exists, skip what doesn't):

Source What to extract
SOUL.md Builder personality, working style, specialisation signals
AGENTS.md Agent architecture sophistication, multi-agent patterns
USER.md Name, location, professional context
MEMORY.md Notable projects, decisions, recurring themes
TOOLS.md Tool integrations, infrastructure knowledge
skills/*/SKILL.md Installed/authored skills (count and categories)
memory/*.md Recent activity, project history, achievements
Conversation exports (if any *.json exports exist in workspace) Real usage patterns, problem-solving depth

Privacy: Never include raw personal data, API keys, secrets, or private conversation content in the profile. Only extract professional capability signals.

2. Build the Profile Object

Construct a JSON object with these fields:

{
  "dn_v": "1.2",
  "issued": "\x3CISO-8601 timestamp>",
  "builder": {
    "name": "\x3Cbuilder name>",
    "location": "\x3Ccity, country — or 'Remote'>",
    "tagline": "\x3Cone-line expertise summary, max 80 chars>"
  },
  "categories": ["\x3C1-3 categories from: Legal, E-commerce, HR, Marketing, Healthcare, Finance, Education, DevOps, Automation, Creative, Data, Security, Research>"],
  "achievements": [
    "\x3Cachievement 1 — specific, quantified where possible>",
    "\x3Cachievement 2>",
    "\x3Cachievement 3>"
  ],
  "skills": ["\x3Cskill 1>", "\x3Cskill 2>", "...up to 8"],
  "systemSummary": "\x3C2-3 sentence summary of builder's agent architecture and approach>",
  "_nonce": "\x3Crandom 8-char alphanumeric string>"
}

Achievement guidelines:

  • Lead with impact ("Reduced X by Y%" / "Built Z handling N items")
  • Be specific — vague claims like "good at prompting" are worthless
  • Derive from actual workspace evidence, not imagination
  • 3 achievements exactly

Category selection:

  • Pick 1-3 categories that best match the builder's demonstrated work
  • If work doesn't fit listed categories, use the closest match

3. Interactive Review

Present the draft profile to the builder before signing:

Here's your Dragnet profile draft:

Name: \x3Cname>
Tagline: \x3Ctagline>
Categories: \x3Ccategories>
Achievements:
  1. \x3Cachievement 1>
  2. \x3Cachievement 2>
  3. \x3Cachievement 3>
Skills: \x3Cskills list>
Summary: \x3CsystemSummary>

Want to adjust anything before I sign and generate the file?

Let the builder edit any field. Re-present after changes until they confirm.

4. Sign and Output

Once confirmed, write the profile JSON to a temp file, then sign it:

# Write unsigned profile
cat > /tmp/dragnet-unsigned.json \x3C\x3C 'PROFILE'
\x3Cthe JSON object without _sig>
PROFILE

# Sign it
node \x3Cskill-dir>/scripts/sign-profile.mjs /tmp/dragnet-unsigned.json ./dragnet-profile.json

The signing script:

  • Computes HMAC-SHA-256 over the canonical JSON (sorted keys, _sig excluded)
  • Appends _sig field
  • Verifies round-trip integrity
  • Writes final dragnet-profile.json

5. Next Steps

After generating, tell the builder:

Your signed profile is at ./dragnet-profile.json.

To go live on Dragnet:

  1. Go to dragnet.unrelated.works
  2. Click Get Listed
  3. Upload your dragnet-profile.json
  4. Dragnet verifies the signature and publishes your profile

To update your profile later, just run this skill again.

Profile Schema Reference

Field Type Required Notes
dn_v string yes Always "1.2"
issued string yes ISO-8601 timestamp
builder.name string yes Full name or pseudonym
builder.location string yes City, Country or "Remote"
builder.tagline string yes Max 80 characters
categories string[] yes 1-3 from allowed list
achievements string[] yes Exactly 3
skills string[] yes 2-8 skill labels
systemSummary string yes 2-3 sentences
_nonce string yes Random 8-char alphanumeric
_sig string auto Added by signing script
安全使用建议
Do not install or use this skill until you confirm two things with the author or marketplace: (1) where the hard-coded key comes from and whether Dragnet actually accepts signatures produced with that key — embedding a verification key in a public skill is a red flag and could allow forged 'verified' profiles; (2) how the skill prevents leaking personal data when it scans conversation exports and memory files. If you plan to use it, inspect the generated unsigned JSON before signing, and prefer a signing workflow that (a) asks you to provide your own secret via an environment variable or (b) uses a proper public-key signature (private key kept secret by the signer). Avoid uploading any profile that may contain raw personal data or API keys. If you need further help, ask the author to remove the hard-coded KEY and to add explicit sanitization steps and a clear explanation of the trust model for Dragnet verification.
能力评估
Purpose & Capability
The skill's declared purpose (generate and sign a Dragnet profile) matches the provided instructions and the included signing script. However, the signing script embeds a hard-coded HMAC key (KEY = "dn-validation-key-alpha-2025"), which is unexpected: a marketplace verification key should not normally be distributed inside a skill. This raises questions about whether the key is a test key, a shared secret that the marketplace expects, or an attempt to enable forging of verified profiles.
Instruction Scope
The SKILL.md instructs the agent to scan many workspace artifacts (SOUL.md, AGENTS.md, USER.md, MEMORY.md, memory/*.md, conversation exports, installed skills). Those sources can contain personal data, private conversation content, or secrets. While the doc states 'Never include raw personal data, API keys, secrets,' it leaves filtering to the agent and provides no concrete sanitization rules or automated redaction, which is a privacy risk and scope creep relative to a simple profile generator.
Install Mechanism
This is an instruction-only skill with one included script; there is no external install, no downloads, and no packages pulled from remote URLs. That minimizes installation risk.
Credentials
The skill declares no required environment variables or external credentials, but the signer script contains a hard-coded secret HMAC key. Embedding a verification key in the code is disproportionate and suspicious: either the skill exposes Dragnet's secret (bad) or it uses an arbitrary key that may let anyone produce signatures accepted by the marketplace (also bad). Also, the skill reads many local files which can expose sensitive user data if not carefully filtered.
Persistence & Privilege
The skill is not always-enabled and does not request special persistent privileges or modify other skills or global agent config. Autonomous invocation is allowed by platform defaults but is not combined with other elevated privileges in this package.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install dragnet
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /dragnet 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release — scan workspace, build verified builder profile, sign with HMAC-SHA-256 for dragnet.unrelated.works
元数据
Slug dragnet
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Dragnet 是什么?

Generate a verified Dragnet marketplace profile. Scans your OpenClaw workspace — SOUL.md, AGENTS.md, USER.md, MEMORY.md, conversation exports, installed skil... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 81 次。

如何安装 Dragnet?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install dragnet」即可一键安装,无需额外配置。

Dragnet 是免费的吗?

是的,Dragnet 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Dragnet 支持哪些平台?

Dragnet 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Dragnet?

由 UnrelatedWorks(@unrelatedworks)开发并维护,当前版本 v1.0.0。

💬 留言讨论