← Back to Skills Marketplace
89
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install gemini-deep-research-jirboy
Description
Performs complex, multi-step research on specified topics, synthesizing web data into detailed markdown reports with progress updates, now integrated into re...
README (SKILL.md)
⚠️ 已整合 - 请使用 research 统一入口
本技能保留用于向后兼容，功能已整合到
research统一入口技能推荐使用：
research deep [主题]或直接使用本技能（自动转发）
Gemini Deep Research（兼容层）
Use Gemini's Deep Research Agent to perform complex, long-running context gathering and synthesis tasks.
迁移指南
新用法：
research deep 竞争分析 电动汽车电池
research deep 市场调研 人工智能
旧用法（仍然可用）：
deep 竞争分析 电动汽车电池
Prerequisites
GEMINI_API_KEYenvironment variable (from Google AI Studio)- Note: Does NOT work with Antigravity OAuth tokens
How It Works
- Breaks down complex queries into sub-questions
- Searches the web systematically
- Synthesizes findings into comprehensive reports
- Provides streaming progress updates
Output
deep-research-YYYY-MM-DD-HH-MM-SS.md- Final report in markdowndeep-research-YYYY-MM-DD-HH-MM-SS.json- Full interaction metadata
Usage Guidance
This skill appears to implement what it claims (a Gemini deep-research client) and only needs your GEMINI_API_KEY and network access to Google's generativelanguage API, but there are several inconsistencies you should resolve or consider before installing:
- The SKILL.md and the Python script require GEMINI_API_KEY, but the registry metadata incorrectly lists no required env vars/credentials. Expect to provide your GEMINI_API_KEY (or run with --api-key).
- The Python script uses the third-party 'requests' library, but no dependency/install instructions are provided. Ensure the runtime has 'requests' available or add an install step.
- Metadata files (_meta.json, package.json) and the registry record show mismatched owner, slug, and version values. That reduces confidence in provenance—ask the publisher to clarify and fix metadata.
- The script performs network calls to https://generativelanguage.googleapis.com and will save report and JSON files locally. If you plan to run it, run in a controlled environment, verify the code yourself, and do not supply sensitive credentials unless you trust the publisher and need the functionality.
If the publisher can fix the metadata and declare the GEMINI_API_KEY requirement and the dependency on 'requests', the skill would be coherent with its purpose. If you cannot confirm provenance, prefer running the script manually in an isolated environment rather than enabling it for autonomous agent invocation.
Capability Analysis
Type: OpenClaw Skill
Name: gemini-deep-research-jirboy
Version: 2.0.0
The skill is a legitimate API client for Google's Gemini Deep Research service. The Python script (scripts/deep_research.py) interacts with the official Google Generative Language API (generativelanguage.googleapis.com) to perform research tasks and save results locally in Markdown and JSON formats, which is consistent with its stated purpose in SKILL.md.
Capability Tags
Capability Assessment
Purpose & Capability
The name/description, SKILL.md, and scripts/deep_research.py consistently implement a Gemini Deep Research client using Google Generative Language API (agent model 'deep-research-pro-preview-12-2025'). However the registry metadata claims no required environment variables or credentials while the SKILL.md and the script both require GEMINI_API_KEY — this is an inconsistency. Additionally the repository metadata in _meta.json (version/owner/slug) does not match the registry metadata (different ownerId/slug/version), which weakens provenance.
Instruction Scope
Runtime instructions and the Python script are scoped to starting and polling a research interaction with the Google API, extracting the report, and saving it to local files. The SKILL.md does not ask the agent to read unrelated files or exfiltrate unrelated credentials. The script writes outputs to the specified output directory and optionally uses a file_search_store tool parameter (which is passed to the API) — nothing in the instructions indicates out-of-scope system access.
Install Mechanism
There is no install spec (instruction-only), which is lower risk. But the shipped Python script imports 'requests' while package.json contains no dependencies and the skill metadata lists no required binaries or packages. That mismatch means the skill may fail at runtime or require ad-hoc installs; the lack of declared dependencies is an oversight that should be corrected.
Credentials
The SKILL.md and script require a GEMINI_API_KEY (or --api-key) to call Google’s generativelanguage API — this is appropriate for the stated purpose. However the registry metadata declares 'required env vars: none' and 'Primary credential: none', which is incorrect and potentially misleading. No other credentials are requested by the code, so the scope of secrets requested is otherwise proportionate, but the metadata omission is a notable risk (users might not realize they need to supply an API key).
Persistence & Privilege
The skill is not marked 'always: true' and is user-invocable only. It does not request to modify other skills or system-wide settings. It writes output files to an output directory (normal behavior). Autonomous invocation remains allowed by platform default but is not a unique privilege of this skill.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install gemini-deep-research-jirboy - After installation, invoke the skill by name or use
/gemini-deep-research-jirboy - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.0.0
Version 2.0.0
- No changes detected in files or documentation.
- Skill behavior and migration notice remain the same.
- Continues to serve as a compatibility layer for the unified research skill.
v1.0.0
- Gemini Deep Research is now merged into the unified research skill entry point.
- This skill remains for backward compatibility; all features are accessible via research.
- Recommended use: research deep [topic] (automatically redirects).
- Legacy usage deep [topic] is still supported.
- No functional changes; update aims to streamline access and migration guidance.
Metadata
Frequently Asked Questions
What is Gemini Deep Research?
Performs complex, multi-step research on specified topics, synthesizing web data into detailed markdown reports with progress updates, now integrated into re... It is an AI Agent Skill for Claude Code / OpenClaw, with 89 downloads so far.
How do I install Gemini Deep Research?
Run "/install gemini-deep-research-jirboy" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Gemini Deep Research free?
Yes, Gemini Deep Research is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Gemini Deep Research support?
Gemini Deep Research is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Gemini Deep Research?
It is built and maintained by JIRBOY (@jirboy); the current version is v2.0.0.
More Skills