← 返回 Skills 市场
ideaforceai-sys

agentcadia-tools

作者 ideaforceai-sys · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
79
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install agentcadia-tools
功能描述
Upload or download Agentcadia agent workspaces with metadata writeback and detailed reporting using explicit upload and download commands.
使用说明 (SKILL.md)

Agentcadia Tools

A unified Agentcadia skill for uploading and downloading agent workspaces.

Core rule

This skill exposes two explicit command-style actions:

  • upload
  • download

Always decide which action the user needs, then run the matching subcommand through the unified script:

python3 scripts/agentcadia_tools.py upload ...
python3 scripts/agentcadia_tools.py download ...

Do not invent alternative flows when the bundled scripts already cover the task.


Action: upload

Use upload when the user wants to sync the current local workspace into an Agentcadia draft.

Required workflow

A successful upload must include all of these phases:

  1. upload workspace markdown + packaged skills
  2. generate draft metadata from the uploaded agent workspace
  3. write that metadata back to Agentcadia
  4. send the owner a standalone metadata-summary message containing the exact metadata that was written back
  5. when a share image is available, send that share image to the owner as a separate image message

If you skip metadata writeback and only provide “suggested copy”, the task is not complete.

Metadata synthesis rules

Generate metadata from the workspace itself.

Prefer agent-level sources such as:

  • AGENTS.md, SOUL.md, IDENTITY.md, USER.md, and TOOLS.md (combine these to infer the agent’s name, personality, self-introduction, intended audience, and operating context)
  • README-like files
  • persona / workflow / intro files
  • other root markdown that describes the whole workspace

Quality bar:

  • summary must read like a short product intro for the agent
  • detailDescription must expand on positioning, scenarios, and value
  • do not write generic template copy
  • do not write “uploaded workspace”, “packaged skills”, or similar process-language as the user-facing description
  • explicitly synthesize from AGENTS.md, SOUL.md, IDENTITY.md, USER.md, and TOOLS.md
  • do not block on agent.md, AGENT.md, or agent_download.md
  • do not ask the owner to provide title/summary/detailDescription/category/tags first unless the workspace is genuinely too sparse; even then, present your best draft first

Required metadata fields:

  • title
  • summary
  • detailDescription
  • description
  • category
  • tags

Upload command

Generate metadata first, save it as a temporary JSON file, then run:

python3 scripts/agentcadia_tools.py upload \
  --task-id \x3CtaskId> \
  --origin \x3Corigin> \
  --workspace \x3CworkspacePath> \
  --metadata-file \x3CtempMetadataJsonPath>

Treat uploader output as authoritative.

If the script returns:

  • success: false
  • or metadataWritebackSucceeded: false
  • or incompleteReason: METADATA_WRITEBACK_REQUIRED

then the upload is not complete, even if files were uploaded.

Owner delivery rules for upload

After upload succeeds, deliver owner-facing output in two separate steps:

  1. Metadata summary message first

    • Send a standalone text message containing at least: title, category, tags, summary, detailDescription, workspace path used, uploaded markdown files, and uploaded skill packages.
    • Do not merge this into the share-image caption only. The owner must be able to read the written metadata as a separate message.
  2. Share image message second

    • If a share image is available, send it as a real image message to the owner.
    • Do not stop at sending only agentUrl or shareImageUrl as plain text.
    • If the runtime can send images only from local files, download the share image first, save it to a temporary local path, then send that local image file.
    • The final owner-visible step should be the image message, not a plain text link.

Upload completion rules

  • upload ok + metadata writeback ok + metadata summary message sent + share image sent as an image when available = complete
  • upload ok + metadata writeback failed = partial success, not complete
  • upload ok + metadata ok + metadata summary not sent = partial success, not complete
  • upload ok + metadata ok + share image available but not sent as an image = partial success, not complete
  • upload failed = failed

Action: download

Use download when the user wants to pull an Agentcadia agent package into the local workspace.

Download behavior

The download flow should:

  1. determine the target workspace path
  2. download the agent package with the provided bearer token
  3. extract or place the files into the workspace
  4. report all placed files and any conflicts
  5. clean up temporary files

Do not overwrite existing files unless the user explicitly asked for it.

Download command

python3 scripts/agentcadia_tools.py download \
  --agent-id \x3CagentId> \
  --token \x3CdownloadToken> \
  --origin \x3Corigin> \
  --workspace \x3CworkspacePath>

If the user explicitly approved overwriting existing files, add:

--allow-overwrite

Download reporting rules

Always report:

  • workspace path used
  • whether the payload was zip or markdown
  • files placed or extracted
  • conflicts skipped because overwrite was not allowed

If any conflicts were skipped, treat the result as partial and tell the user exactly which files need a decision.


Runtime notes

When upload returns a share image, sending that image to the owner is mandatory unless the runtime truly cannot send images. The owner-facing metadata summary and the share image must be sent as two separate messages, with the image sent last.

See references/openclaw-runtime.md for OpenClaw-oriented owner-delivery behavior.

Boundaries

  • Use this skill for deterministic file sync, not final publish approval
  • Never expose upload tokens or internal upload URLs in user-visible output
  • Stop and report clearly if a packaged skill exceeds size limits
  • Do not fabricate share assets if the server did not return them
  • Do not overwrite existing local files during download unless the user explicitly approved it
安全使用建议
This skill appears coherent and implements the claimed upload/download flows. Before installing or running it: 1) Only use with a trusted Agentcadia origin and task/token you obtained from a trusted workflow. 2) Treat upload/download tokens as secrets — do not paste them into public chat or logs; verify the final JSON output for any accidental token exposure before sending it to others. 3) The skill will read and package any files in the chosen workspace — remove any secrets or credentials from the workspace before uploading. 4) The runtime (host) is responsible for delivering the owner-facing messages and images exactly as SKILL.md requires; verify your runtime won't unintentionally leak internal upload URLs or tokens when forwarding script responses. 5) The download path fallback will create ~/.openclaw/workspace if no workspace is found; confirm that behavior is acceptable. If you want extra assurance, review the remainder of upload_agentcadia.py (truncated here) to confirm it does not print or return sensitive tokens or internal URLs in its final output.
功能分析
Type: OpenClaw Skill Name: agentcadia-tools Version: 1.0.0 The skill bundle facilitates the automated packaging and transmission of local workspace files (markdown and skill ZIPs) to a remote endpoint via `upload_agentcadia.py`. While this aligns with the stated purpose of syncing with the 'Agentcadia' platform, the tool's design allows for data transfer to any user-provided URL via the `--origin` flag. The `SKILL.md` instructions direct the AI to synthesize and exfiltrate metadata from sensitive local files like `SOUL.md` and `IDENTITY.md`. Although `download_agentcadia.py` includes a ZipSlip mitigation (checking for '..' in paths), the core functionality provides a high-risk mechanism for data exfiltration and remote file execution if misconfigured or targeted at sensitive directories.
能力标签
requires-oauth-token
能力评估
Purpose & Capability
Name/description match what the code does: the included scripts implement upload and download flows against a user-supplied Agentcadia origin and tokens. There are no unrelated required env vars or binaries and the CLI args (taskId, token, origin, workspace, metadata) are appropriate for the stated functionality.
Instruction Scope
SKILL.md instructs the agent to synthesize metadata from workspace files and to call the bundled scripts; the scripts accept metadata and workspace paths and perform only the expected actions (HTTP calls to the provided origin, zipping skill dirs, extracting downloads, writing files). The runtime must implement owner messaging (metadata text + image) as SKILL.md requires — the scripts emit JSON results but do not perform messaging themselves. The scripts only read files inside the chosen workspace and temporary directories.
Install Mechanism
This is instruction-only (no install spec). All code is bundled in the skill; nothing is downloaded or written to disk beyond normal temporary files created at runtime. No external install URLs or archive extraction during install.
Credentials
The skill requests no environment variables or system-level credentials. Sensitive values (upload/download tokens) are passed as CLI arguments by design, which is proportional to the task. The scripts read workspace files (expected for packaging/upload) — ensure those files do not contain secrets you don't want uploaded.
Persistence & Privilege
The skill does not request permanent or always-on presence (always: false). It does not modify other skills or system-wide settings; it uses ephemeral temp directories for packaging and extraction.
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 agentcadia-tools
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /agentcadia-tools 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 of agentcadia-tools: unified upload and download workflows for Agentcadia agent workspaces. - Provides two main commands: `upload` (sync local workspace to Agentcadia) and `download` (fetch Agentcadia agent package locally). - Enforces full metadata writeback and reporting for uploads, including synthesized agent metadata and owner-facing summaries. - Requires explicit user approval before overwriting files on download. - Mandates that share images are delivered as separate image messages when available. - Offers deterministic, script-driven file transfer and metadata management for Agentcadia automation tasks.
元数据
Slug agentcadia-tools
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

agentcadia-tools 是什么?

Upload or download Agentcadia agent workspaces with metadata writeback and detailed reporting using explicit upload and download commands. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 79 次。

如何安装 agentcadia-tools?

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

agentcadia-tools 是免费的吗?

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

agentcadia-tools 支持哪些平台?

agentcadia-tools 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 agentcadia-tools?

由 ideaforceai-sys(@ideaforceai-sys)开发并维护,当前版本 v1.0.0。

💬 留言讨论