> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mira-app.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Built-in Skills

> The five Skills that ship with MIRA — domain-expert reasoning personas ready to apply to any session.

<Note>
  Skills are **fully functional** and ready to use. Since built-in skills are indicative, feel free
  to tune them as needed while we continuously roll out new built-in skills, configuration options,
  and quality improvements.
</Note>

MIRA ships with five built-in Skills covering the most common professional domains. Each Skill configures the reasoning engine with domain-specific instructions, a tuned temperature, and iteration limits appropriate for that domain's requirements.

## What a Skill does

When a Skill is active in a session, the reasoning engine:

* Receives the Skill's instruction set as part of its system context
* Applies the Skill's model overrides (temperature, max iterations, max tokens)
* Operates within the Skill's MCP tool allowlist
* Maintains the Skill's persona for the full session duration

## Built-in Skills

<Tabs>
  <Tab title="🔬 Life Sciences Expert">
    **Category:** Domain Expert\
    **Icon:** 🔬\
    **Colour:** Emerald

    **Optimised for:**

    * Pharmaceutical research and drug development pipelines
    * Clinical trial design, ICH/GCP guidelines, FDA 21 CFR compliance
    * Bioinformatics: sequence analysis, NGS pipelines, genomics
    * Statistical analysis for biomedical data (survival analysis, dose-response)
    * Regulatory frameworks: EMA, FDA, ICH Q1–Q14 guidelines

    **Preferred libraries:** biopython, pandas, scipy, statsmodels, lifelines

    **Model overrides:**

    | Override       | Value |
    | -------------- | ----- |
    | Temperature    | 0.3   |
    | Max iterations | 20    |
    | Max tokens     | 8,192 |
  </Tab>

  <Tab title="💻 Coding & Architecture">
    **Category:** Coding\
    **Icon:** 💻\
    **Colour:** Blue

    **Optimised for:**

    * Writing clean, idiomatic, well-documented code in any language
    * Applying SOLID principles and appropriate design patterns
    * Code review with actionable, specific feedback
    * System design: scalability, fault tolerance, API design
    * Performance optimisation and profiling
    * Security best practices: input validation, auth patterns, secrets management

    **Model overrides:**

    | Override       | Value |
    | -------------- | ----- |
    | Temperature    | 0.2   |
    | Max iterations | 25    |
    | Max tokens     | 8,192 |
  </Tab>

  <Tab title="📊 Financial Analyst">
    **Category:** Domain Expert\
    **Icon:** 📊\
    **Colour:** Amber

    **Optimised for:**

    * Financial statement analysis (P\&L, balance sheet, cash flow)
    * Valuation models: DCF, comparables, precedent transactions
    * Risk analysis: VaR, CVaR, Monte Carlo simulation
    * Time series analysis for financial data
    * Portfolio construction and optimisation (Markowitz, Black-Litterman)
    * Derivatives pricing and hedging strategies

    **Preferred libraries:** pandas, numpy, scipy, yfinance, quantlib

    **Model overrides:**

    | Override       | Value |
    | -------------- | ----- |
    | Temperature    | 0.2   |
    | Max iterations | 20    |
    | Max tokens     | 8,192 |
  </Tab>

  <Tab title="🔒 Security Reviewer">
    **Category:** Coding\
    **Icon:** 🔒\
    **Colour:** Red

    **Optimised for:**

    * Application security: OWASP Top 10, injection attacks, XSS, CSRF
    * Cryptography: proper algorithm selection, key management, TLS configuration
    * Authentication and authorisation patterns (OAuth 2.0, JWT, RBAC)
    * Infrastructure security: network segmentation, secrets management, IAM
    * Threat modelling: STRIDE, attack trees, data flow analysis
    * Secure code review with CVE awareness

    **Findings format:** HIGH / MEDIUM / LOW severity with specific remediation code and OWASP/NIST/CWE references.

    **Model overrides:**

    | Override       | Value |
    | -------------- | ----- |
    | Temperature    | 0.1   |
    | Max iterations | 20    |
    | Max tokens     | 8,192 |
  </Tab>

  <Tab title="🎓 Research Synthesizer">
    **Category:** Research\
    **Icon:** 🎓\
    **Colour:** Purple

    **Optimised for:**

    * Systematic literature review and meta-analysis methodology
    * Synthesising findings across multiple sources with attribution
    * Identifying research gaps, contradictions, and emerging trends
    * Structuring research outputs: abstracts, literature reviews, discussion sections
    * Citation management and academic formatting (APA, Vancouver, IEEE)
    * Critical appraisal of study design and statistical methods

    **Output format:** Background → Methods → Results → Discussion → Conclusion

    **Model overrides:**

    | Override       | Value |
    | -------------- | ----- |
    | Temperature    | 0.4   |
    | Max iterations | 25    |
    | Max tokens     | 8,192 |
  </Tab>
</Tabs>

## Applying a built-in skill

1. Click **New Chat** (⌘N)
2. In the Skill Picker, click the skill you want
3. The skill badge appears in the session header
4. Start your query — the engine operates with the skill's persona immediately

You can also apply a skill mid-session: click the skill badge in the session header → **Change Skill** → select a new skill.

## Related

<CardGroup cols={2}>
  <Card title="Creating Custom Skills" icon="plus" href="/features/skills/creating-custom-skills">
    Build your own domain-specific reasoning persona
  </Card>

  <Card title="Skill Scope" icon="globe" href="/features/skills/skill-scope">
    Apply a skill to one session or all new sessions automatically
  </Card>
</CardGroup>