Tech Analyzer

Deep analysis of technology researchers/innovators and their core technical advantages, with comparative analysis of similar teams domestically and internati...

Security Scan

Skill mostly does what it says (generate comparative technical reports) but the runtime instructions reference external tools/binaries (pandoc, kimi_fetch/kimi_search/web_fetch/read) that are not declared in the metadata — this mismatch deserves clarification before installing.

Purpose & Capability

The skill's declared purpose (analyze researchers and produce a DOCX report) aligns with the included templates and the helper script. However, SKILL.md explicitly requires pandoc for DOCX conversion and references platform tools (kimi_fetch, kimi_search, web_fetch, read) while the skill metadata declares no required binaries or environment. The missing declaration for pandoc and explicit tool dependencies is an inconsistency the author should justify or fix.

Instruction Scope

Runtime instructions are specific about inputs (PDF/DOCX/URLs/text), extraction methods, and search queries. They instruct the agent to fetch external URLs and perform broad searches (domestic and international). This is expected for the skill's purpose, but it grants the agent broad discretion to access arbitrary web resources and aggregate content — the SKILL.md does not add limits or indicate privacy/consent handling for sensitive files.

Install Mechanism

There is no install spec (lowest-risk delivery). The included Python script is simple and transparent. The main concern is an unstated dependency: SKILL.md instructs use of pandoc (a binary) for DOCX generation but the skill metadata does not list pandoc as required nor provide an install step or fallback. That omission should be addressed.

Credentials

The skill requests no environment variables, credentials, or config paths. The scope of requested data is limited to user-provided documents/URLs and web search results, which is proportionate to the stated purpose.

Persistence & Privilege

The skill is not always-enabled and does not request persistent privileges or modify other skills. It relies on on-demand invocation only (default behavior).

What to consider before installing

Before installing: 1) Confirm whether pandoc is available in the environment or ask the author to add a fallback (e.g., python-docx) or declare pandoc in required binaries/install steps. 2) Verify that the platform tools referenced (kimi_fetch, kimi_search, web_fetch, read) are available and permitted to be used by this skill — ask the author to list required platform tools explicitly. 3) Be aware the skill will fetch arbitrary URLs and ingest uploaded documents (CVs, papers). Don’t upload sensitive or proprietary material unless you trust the environment and know how fetched content is stored/retained. 4) If you need tighter control, request the SKILL.md be updated to: (a) declare all external tool/binary dependencies, (b) limit search domains or include opt-in for web searching, and (c) state data retention/where outputs are stored. These clarifications would move this assessment toward benign.

Tech Analyzer Skill

Analyze technology researchers/innovators and their competitive landscape.

Workflow

Step 1: Input Processing

Accept input in any of these forms:

• File attachments: PDF, DOCX, or other document formats
• Web links: URLs to articles, profiles, or reports
• Direct text: Pasted content about the target person/technology

Use appropriate tools to extract content:

• For files: Read file content
• For URLs: Use kimi_fetch or web_fetch to extract content
• For text: Use directly

Step 2: Content Analysis

Analyze the extracted content to identify:

Person Profile:

• Name, title, affiliation
• Research field and focus area
• Key achievements and recognitions

Core Technical Advantages:

• Breakthrough innovations
• Key methodologies or systems
• Performance metrics and validation results
• Unique insights or theoretical contributions
• Academic and industrial impact

Step 3: Competitive Research

Search for similar teams/researchers both domestically (China) and internationally:

Domestic Search (China):

• Leading universities (Tsinghua, Peking University, Fudan, etc.)
• Research institutes (CAS institutes)
• Industry companies and startups
• Key research groups in the same field

International Search:

• Top universities (MIT, Stanford, CMU, etc.)
• Research labs (Google Research, OpenAI, etc.)
• Industry leaders and startups
• Academic conferences and publications

Use kimi_search with specific queries to find relevant information.

Step 4: Comparative Analysis

Create a structured comparison covering:

Dimension	Target Person	Domestic Peers	International Peers
Approach	[Methodology]	[Domestic approach]	[International approach]
Advantages	[Key strengths]	[Domestic strengths]	[International strengths]
Stage	[Research/validation/product]	[Domestic stage]	[International stage]

Step 5: Report Generation

Generate a comprehensive Word (.docx) document with the following structure:

# [Person Name] Technical Analysis Report

## 1. Person Profile
- Basic information
- Affiliation and title
- Key achievements

## 2. Core Technical Advantages

### 2.1 Breakthrough Innovations
[Description of key systems/methods]

### 2.2 Key Methodologies
[Technical approach and unique insights]

### 2.3 Performance Validation
[Metrics, benchmarks, validation results]

### 2.4 Academic & Industrial Impact
[Publications, citations, technology transfer]

## 3. Competitive Landscape Analysis

### 3.1 Domestic Teams (China)
| Organization | Research Focus | Key Progress |
|--------------|----------------|--------------|
| [Team 1] | [Focus] | [Progress] |
| [Team 2] | [Focus] | [Progress] |

### 3.2 International Teams
| Organization | Research Focus | Key Progress |
|--------------|----------------|--------------|
| [Team 1] | [Focus] | [Progress] |
| [Team 2] | [Focus] | [Progress] |

## 4. Comparative Summary

### 4.1 Target Person's Uniqueness
[List unique advantages]

### 4.2 Domestic vs International Landscape
[Overall comparison and gap analysis]

### 4.3 Technology Trajectory
[Future directions and trends]

---
*Report generated: [Date]*
*Sources: [List key sources]*

Use pandoc to convert Markdown to DOCX:

pandoc report.md -o report.docx

Step 6: Output Delivery

1. Save the .docx file to workspace
2. Send file to user via message tool with filePath parameter
3. If file transfer fails, provide the content directly as formatted text

Research Guidelines

Search Strategy:

• Use kimi_search with include_content=true for comprehensive results
• Search for both academic (universities, institutes) and industry (companies, startups) perspectives
• Include key terms in both Chinese and English for broader coverage

Analysis Principles:

• Be specific: Include names, institutions, system names, metrics
• Be comparative: Highlight differences in approaches
• Be balanced: Acknowledge both strengths and limitations
• Be current: Focus on recent developments (last 2-3 years)

Quality Standards:

• Technical accuracy: Verify technical claims from reliable sources
• Completeness: Cover all required sections
• Objectivity: Present balanced view of competitive landscape
• Actionability: Provide insights useful for decision-making

Error Handling

If input cannot be processed:

1. Ask user for clarification or alternative format
2. If content extraction fails, request user to paste text directly

If research yields limited results:

1. Expand search terms
2. Try related fields or technologies
3. Note limitations in the report

If file generation fails:

1. Provide content as formatted Markdown
2. Offer to retry with different approach