Skip to content

Troubleshooting

Common issues and solutions for the Startup Valuation library, MCP server, and AI Agent Skills.

Installation Issues

fastmcp not found

Error: ModuleNotFoundError: No module named 'fastmcp'

Solution: Install the MCP extras:

pip install startup-valuation[mcp]

startup_valuation not found

Error: ModuleNotFoundError: No module named 'startup_valuation'

Solution: Install the library:

pip install startup-valuation

Version mismatch

Error: Unexpected behavior or missing functions

Solution: Check your installed version:

pip show startup-valuation

Ensure you're on the latest version:

pip install --upgrade startup-valuation

MCP Server Issues

Connection refused

Error: Connection refused or ECONNREFUSED

Causes and solutions:

  1. Server not running — Start the server first:

    cd mcp_server && python server.py
    

  2. Wrong transport mode — stdio mode doesn't use network ports. If you're trying to connect via HTTP, configure SSE transport:

    mcp.run(transport="sse", host="0.0.0.0", port=8000)
    

  3. Firewall blocking — Ensure port 8000 (or your configured port) is not blocked.

Tool not found

Error: Tool "valuation_xyz" not found

Solutions:

  1. Check tool name spelling — Tool names use valuation_ prefix and snake_case:
  2. valuation_scorecard
  3. valuation_score_card
  4. scorecard_valuation (this is the library function name, not the MCP tool name)

  5. Check the tool reference for the complete list of 60+ tools.

Server crashes on startup

Error: Python traceback on python server.py

Solutions:

  1. Check Python version — Requires Python 3.10+:

    python --version
    

  2. Check dependencies — Reinstall with MCP extras:

    pip install --force-reinstall startup-valuation[mcp]
    

  3. Check for conflicting packages — If you have multiple versions of fastmcp:

    pip uninstall fastmcp
    pip install startup-valuation[mcp]
    

AI Agent Skills Issues

Skills not loading

Error: Agent doesn't recognize the skills

Solutions:

  1. Check file path — For OpenCode, skills must be in ~/.config/opencode/skills/:

    ls ~/.config/opencode/skills/
    # Should show: valuation-core  valuation-advanced  valuation-industry  valuation-stakeholder  valuation-emerging
    

  2. Check file name — Each skill directory must contain a SKILL.md file:

    ls ~/.config/opencode/skills/valuation-core/
    # Should show: SKILL.md
    

  3. Check SKILL.md format — The file must start with # Skill: header:

    head -1 ~/.config/opencode/skills/valuation-core/SKILL.md
    # Should show: # Skill: valuation-core
    

  4. Restart the agent — Skills are loaded at startup. Restart OpenCode/Claude/Cursor after adding skills.

Agent doesn't follow skill workflow

Issue: Agent ignores the skill instructions

Solutions:

  1. Check skill is in the right location — Different agents use different skill paths:
  2. OpenCode: ~/.config/opencode/skills/
  3. Claude: Custom Instructions (paste SKILL.md content)
  4. Cursor: Custom Instructions (paste SKILL.md content)

  5. Make the skill name specific — Rename the skill to match user queries:

  6. valuation-saas (matches "value my SaaS startup")
  7. valuation-industry (too generic)

  8. Add trigger keywords — In the When to Use section, include phrases users commonly say:

    ## When to Use
    - User says "value my SaaS" or "SaaS valuation"
    - User mentions ARR, MRR, churn, or LTV
    

Python Library Issues

Unexpected results

Issue: Function returns unexpected valuation

Solutions:

  1. Check parameter units — All monetary values are in absolute dollars (not thousands or millions):

    # Correct: $1.5 million
    scorecard_valuation(average_valuation=1_500_000, ...)
    
    # Wrong: passing 1.5 (interpreted as $1.50)
    scorecard_valuation(average_valuation=1.5, ...)
    

  2. Check parameter constraints — Some functions validate inputs:

  3. scorecard_valuation: weights must sum to 1.0
  4. berkus_valuation: each factor must be 0 to max_per_factor
  5. risk_factor_summation: exactly 12 risk ratings, each -2 to +2

  6. Check the API reference for parameter details and constraints.

Import errors

Error: ImportError: cannot import name 'xyz' from 'startup_valuation'

Solution: Check the correct module:

# Core methods
from startup_valuation.core import scorecard_valuation, berkus_valuation

# Advanced methods
from startup_valuation.advanced import black_scholes, monte_carlo_valuation

# Industry-specific
from startup_valuation.saas import ltv_saas, cac
from startup_valuation.biotech import rnPV, decision_tree_ev

# All types
from startup_valuation.types import ValuationResult, Scenario, Distribution

See the chapter index for the full module-to-function mapping.