AI-Agent Skills¶
The Intangible Valuation platform includes AI-Agent Skills that provide domain-specific instructions and workflows for valuation tasks.
Overview¶
Skills are specialized instruction sets that guide AI agents through complex valuation workflows. Each skill contains:
- Domain-specific knowledge and best practices
- Step-by-step workflow guidance
- Method selection criteria
- Common pitfalls and how to avoid them
- Reference to relevant textbook chapters
Available Skills¶
Asset Valuation Skill¶
Guides AI agents through the complete valuation process for different intangible asset types:
- Patents — Relief from Royalty, incremental cash flow, litigation damages
- Trademarks/Brands — Relief from Royalty, royalty benchmarking, brand strength analysis
- Technology/Software — Replacement cost, multi-period excess earnings, useful life estimation
- Customer Relationships — Multi-period excess earnings, attrition analysis
- Human Capital — Cost approach, excess earnings, replacement cost
Discount Rate Construction Skill¶
Walks through building appropriate discount rates:
- Select method (build-up, CAPM, WACC) based on asset type and data availability
- Gather risk-free rate, equity risk premium, size premium
- Add industry-specific and company-specific risk premiums
- Apply adjustments (DLOM, control premium, currency risk)
Purchase Price Allocation Skill¶
Step-by-step PPA workflow:
- Determine total purchase consideration
- Identify and value tangible assets at fair value
- Identify intangible assets and select valuation methods
- Value each identified intangible
- Calculate assumed liabilities
- Compute goodwill as residual
- Validate allocation percentages
Monte Carlo Analysis Skill¶
Guide for uncertainty analysis:
- Identify uncertain input parameters
- Select appropriate probability distributions
- Configure iteration count and random seed
- Run simulation and interpret results
- Perform sensitivity analysis on key drivers
Decision Tree Skill¶
For valuing assets with contingent outcomes:
- Map decision points and chance events
- Assign probabilities and values to each branch
- Apply backward induction
- Identify optimal decision path
- Calculate expected value
Using Skills¶
Skills are located in the skills/ directory. Each skill is a markdown file with:
- Trigger conditions — When to use this skill
- Workflow steps — Ordered steps to follow
- Method selection — Decision trees for choosing valuation methods
- Validation checks — Sanity checks on results
- References — Textbook chapter and section references
Integration with MCP Server¶
Skills complement the MCP server tools:
- The skill provides the workflow and decision logic
- The MCP tools execute the calculations
- The skill interprets results and guides next steps
Example workflow:
User: "Value a pharmaceutical patent with 12 years remaining"
Skill triggers: Asset Valuation -> Patent
1. Determine useful life (min of legal and economic)
2. Select valuation method (Relief from Royalty recommended)
3. Call MCP tool: royalty_rate_benchmark("patent", "pharmaceutical")
4. Call MCP tool: relief_from_royalty(...)
5. Validate result against benchmark range
6. Present final valuation with confidence range
Custom Skills¶
To create a custom skill:
- Create a new markdown file in
skills/ - Define trigger conditions
- Write workflow steps with MCP tool calls
- Include validation checks and references
- Test with sample valuations
Skill Best Practices¶
- Always validate inputs before calling valuation tools
- Use benchmark data to sanity-check results
- Document all assumptions made during the valuation
- Provide sensitivity analysis for key inputs
- Cite the relevant textbook chapter for each method used