skills/query_classification/SKILL.md
```markdown # SKILL: Query Classification ## Task Description This skill classifies user queries into predefined categories based on the complexity and required resources to answer them. The goal is to route queries to the appropriate processing pipeline. ## Output Format The output should be a single string representing the category. Possible categories are: `SIMPLE`, `RESEARCH`. ## Decision Framework 1. **Initial Assessment:** Quickly scan the query for keywords and phrases indicating compl
npx skillsauth add prathamchopra001/inquiro skills/query_classificationInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
# SKILL: Query Classification
## Task Description
This skill classifies user queries into predefined categories based on the complexity and required resources to answer them. The goal is to route queries to the appropriate processing pipeline.
## Output Format
The output should be a single string representing the category. Possible categories are: `SIMPLE`, `RESEARCH`.
## Decision Framework
1. **Initial Assessment:** Quickly scan the query for keywords and phrases indicating complexity or research intent.
2. **SIMPLE Classification:**
- **Criteria:** The query can be answered directly from readily available general knowledge. The answer should be concise (1-3 sentences).
- **Keywords:** "What is...", "Who is...", "When did...", "Define..."
- **Confidence:** High confidence if the query fits the criteria and uses simple vocabulary.
3. **RESEARCH Classification:**
- **Criteria:** The query requires investigation, literature review, data analysis, or in-depth exploration to provide a comprehensive answer. It often involves identifying factors, analyzing relationships, or exploring complex topics.
- **Keywords:** "Investigate...", "Analyze...", "Compare...", "What are the factors...", "What is the relationship between..."
- **Confidence:** High confidence if the query necessitates going beyond readily available knowledge.
4. **Edge Case Handling:**
- If a query is ambiguous, lean towards `RESEARCH`. It's better to over-classify complexity than to under-classify.
- If a query has multiple parts, and at least one part requires research, classify as `RESEARCH`.
5. **Confidence Scoring:** Not explicitly scored, but used internally to guide the classification. High confidence means the query clearly aligns with the category's criteria.
## Common Patterns
**Example 1:**
* **Input:** "What is the capital of France?"
* **Output:** `SIMPLE`
**Example 2:**
* **Input:** "Investigate the impact of social media on teenage mental health."
* **Output:** `RESEARCH`
**Example 3:**
* **Input:** "What is the function of mitochondria, and how does mitochondrial dysfunction contribute to neurodegenerative diseases?"
* **Output:** `RESEARCH` (because the second part of the question requires research)
## Anti-Patterns
1. **Misclassifying complex questions as simple:** Avoid classifying queries requiring detailed explanations or multiple sources as `SIMPLE`.
2. **Overclassifying simple questions as research:** Avoid classifying basic factual questions as `RESEARCH` simply because they are related to a complex topic.
3. **Returning invalid categories:** Do not return any categories other than `SIMPLE` or `RESEARCH`.
## Quality Checklist
1. Does the classification align with the complexity of the query?
2. Is the answer to the query readily available general knowledge, or does it require research?
3. Does the classification use only the allowed categories (`SIMPLE`, `RESEARCH`)?
4. If the query has multiple parts, is the most complex part used for classification?
development
# SKILL.md: task_generation ## Task Description Generates focused research tasks to advance a scientific objective by addressing knowledge gaps and balancing exploration/exploitation. Prioritizes tasks that validate strong findings or explore high-impact hypotheses. ## Output Format **JSON Array** of objects with these fields: - `type`: "Exploration" (new hypotheses) or "Exploitation" (existing hypothesis validation) - `description`: Specific action to take (e.g., "Test hypothesis X un
development
```markdown # SKILL: Scoring ## Task Description This role is responsible for assigning scores or ratings based on predefined criteria. It involves evaluating information and applying a consistent scoring rubric. ## Output Format The output should be a JSON object with the following structure: ```json { "score": INTEGER, "reason": STRING, "confidence": FLOAT (0.0 to 1.0) } ``` ## Decision Framework 1. **Identify Criteria:** Understand the specific criteria to be used for scoring. This
development
```markdown # SKILL: schema_design ## Task Description This skill focuses on designing effective and well-structured schemas for various data types, ensuring data integrity and facilitating efficient data processing. It involves defining the structure, data types, and constraints for datasets. ## Output Format Output should be a well-formatted JSON schema definition, including: * `type`: Data type (e.g., "object", "string", "number", "array", "boolean") * `properties`: (For objects) A dict
development
```markdown # SKILL: Report Writing ## Task Description This skill focuses on synthesizing research findings into concise, well-structured reports. It involves identifying key themes, patterns, and insights from individual findings, and presenting them in a coherent and informative manner. ## Output Format The output should be a structured report, formatted as follows: ```json { "report_title": "Title of the Report", "executive_summary": "A brief overview of the report's key findings and