← 返回 Skills 市场
joelchance

Manus AI Agent

作者 joelchance · GitHub ↗ · v1.0.1
cross-platform ⚠ suspicious
3703
总下载
7
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install manus-ai
功能描述
Create and manage AI agent tasks via Manus API. Manus is an autonomous AI agent that can browse the web, use tools, and deliver complete work products.
使用说明 (SKILL.md)

Manus AI Agent

Create tasks for Manus, an autonomous AI agent, and retrieve completed work products.

Authentication

Set MANUS_API_KEY env var with your key from manus.im.

Commands

All commands use scripts/manus.sh.

Create a Task

{baseDir}/scripts/manus.sh create "Your task description here"
{baseDir}/scripts/manus.sh create "Deep research on topic" manus-1.6-max

Profiles: manus-1.6 (default), manus-1.6-lite (fast), manus-1.6-max (thorough).

Check Status

{baseDir}/scripts/manus.sh status \x3Ctask_id>

Returns: pending, running, completed, or failed.

Wait for Completion

{baseDir}/scripts/manus.sh wait \x3Ctask_id>
{baseDir}/scripts/manus.sh wait \x3Ctask_id> 300  # custom timeout in seconds

Polls until task completes or times out (default: 600s).

Get Task Details

{baseDir}/scripts/manus.sh get \x3Ctask_id>

Returns full task JSON including status and output.

List Output Files

{baseDir}/scripts/manus.sh files \x3Ctask_id>

Shows filename and download URL for each output file.

Download Output Files

{baseDir}/scripts/manus.sh download \x3Ctask_id>
{baseDir}/scripts/manus.sh download \x3Ctask_id> ./output-folder

Downloads all output files to the specified directory (default: current directory).

List Tasks

{baseDir}/scripts/manus.sh list

Typical Workflow

  1. Create task: manus.sh create "your prompt"
  2. Wait for completion: manus.sh wait \x3Ctask_id>
  3. Download results: manus.sh download \x3Ctask_id>

Advanced API Features

For file attachments, webhooks, connectors, projects, multi-turn conversations, and interactive mode, see the full Manus API documentation:

Security & Permissions

What this skill does:

  • Sends task prompts to the Manus API at api.manus.ai
  • Polls for task completion and downloads output files from Manus CDN
  • API key is sent only in the API_KEY header to api.manus.ai

What this skill does NOT do:

  • Does not upload local files (file upload is an advanced API feature not implemented in the bundled script)
  • Does not register webhooks or connect external accounts
  • Does not send your API key to any endpoint other than api.manus.ai
  • Does not modify local system configuration
  • Cannot be invoked autonomously by the agent (disable-model-invocation: true)
  • You must explicitly trigger every Manus task

Bundled scripts: scripts/manus.sh (Bash — uses curl and jq)

Review scripts/manus.sh before first use to verify behavior.

安全使用建议
This skill appears to do what it says: it sends prompts to api.manus.ai and downloads result files. Before use: (1) review the bundled scripts (already included) and confirm you are comfortable with the API endpoint and header usage, (2) consider using a Manus API key with minimal scope and rotate/revoke it if needed, (3) be cautious opening downloaded files (they come from URLs returned by the API and may be arbitrary content), and (4) if you need stronger guarantees, test with a non-production API key and check network logs to verify requests go only to Manus domains.
功能分析
Type: OpenClaw Skill Name: manus-ai Version: 1.0.1 The skill is classified as suspicious due to shell injection and path traversal vulnerabilities in `scripts/manus.sh`. Specifically, the `task_id` argument in several commands (e.g., `get`, `status`, `wait`, `files`, `download`) and the `output_dir` argument in the `download` command are directly interpolated into shell commands (`curl`, `mkdir`) without proper sanitization, allowing for arbitrary command execution. Additionally, the `output_dir` is not sanitized for path traversal, enabling files to be written to arbitrary locations. While `SKILL.md` correctly sets `disable-model-invocation: true` (mitigating autonomous agent exploitation) and the `create` command's prompt is properly sanitized with `jq -Rs .`, these direct shell vulnerabilities pose a significant risk if a malicious user invokes the skill.
能力评估
✓ Purpose & Capability
Name/description (create/manage Manus agent tasks) match the requested resources: curl and jq for HTTP/JSON work and a single MANUS_API_KEY. The script talks only to api.manus.ai and the flows (create, poll, list, download) align with an API client.
✓ Instruction Scope
SKILL.md instructs use of the bundled scripts and explicitly documents behavior. The script only reads MANUS_API_KEY, calls the Manus API, polls status, and downloads files referenced in task outputs. It does not read other system files or environment variables, nor does it attempt to upload local files or modify system configuration.
✓ Install Mechanism
Instruction-only skill with a small bundled bash script; no install spec and no external archive downloads. No high-risk install mechanisms observed.
✓ Credentials
Only a single credential (MANUS_API_KEY) is required and used directly to authenticate requests to the Manus API. That is proportional to the client's functionality.
✓ Persistence & Privilege
Skill is not always-on, and model invocation is disabled (agent cannot invoke autonomously). The skill does not attempt to change other skills or system-wide settings.
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 manus-ai
  3. 安装完成后，直接呼叫该 Skill 的名称或使用 /manus-ai 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
v1.0.1
- Version bump to 1.0.1 with no file or documentation changes. - No updates or modifications were detected in this release.
v1.0.0
Initial release of the Manus skill. - Enables creating, managing, and retrieving tasks from the Manus autonomous AI agent via CLI. - Supports task creation, status checks, polling for completion, and downloading results. - Allows selection of Manus agent profiles (default, lite, max). - Requires `MANUS_API_KEY` environment variable and `curl`, `jq` utilities. - Includes clear list of commands and describes access and security details.
元数据
Slug manus-ai
版本 1.0.1
许可证 —
累计安装 1
当前安装数 0
历史版本数 2
常见问题

Manus AI Agent 是什么？

Create and manage AI agent tasks via Manus API. Manus is an autonomous AI agent that can browse the web, use tools, and deliver complete work products. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 3703 次。

如何安装 Manus AI Agent？

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

Manus AI Agent 是免费的吗？

是的，Manus AI Agent 完全免费（开源免费），可自由下载、安装和使用。

Manus AI Agent 支持哪些平台？

Manus AI Agent 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 Manus AI Agent？

由 joelchance（@joelchance）开发并维护，当前版本 v1.0.1。

推荐 Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191115 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 留言讨论