Skip to content

Ken-Technology/cold-email-skills / qualification

qualification

Verified

Write AI prompts that judge whether a prospect fits the campaign's audience: who is in, who is out, and why. Use when cleaning or qualifying a prospect list for cold email. Triggers on "qualify my leads", "who should we exclude", "write a qualification prompt".

FileKindSize
SKILL.mdskill md6.3 KB
references/qualification-guide.mdmarkdown3.1 KB

references/qualification-guide.md

View on GitHub
# Qualification Prompt Guidelines

## Purpose

Write the structured variables an AI qualifier (any LLM-based qualification step) uses to judge each prospect. Three fields build the prompt; they are the three H2 sections of `qualification.md`.

## Structured Fields

| Field | Required | Purpose |
|-------|----------|---------|
| Audience description | Yes | Who the target audience is, in one short sentence |
| Disqualification criteria | Yes (preferred) | Conditions that DISQUALIFY a prospect |
| Qualification criteria | Optional | Conditions that QUALIFY a prospect |

At least 2 of 3 fields must be filled. Strong preference for audience description + disqualification criteria only.

## Key Principles

### Write Short, Trust the AI
- The qualification AI is intelligent - give it a brief and let it reason, don't script every edge case
- A few high-signal lines beat a long rulebook. No exhaustive lists of titles, industries, or product names
- Shorter prompts = more room for the AI's own judgment, which is the point

### Never Repeat Search Filters
- `search-strategy.md` already bounds the list. Whatever dimension it sets - headcount, geography, title, seniority, industry - the prompt says NOTHING about it
- Example: if the filters set company headcount to 2,000-50,000, do not mention employee count anywhere in the prompt. If the filters set geography to US/Canada, don't name a country. The list is already that.
- Restating a filter wastes the AI's attention and can fight the filter. Qualification only catches what the filters CAN'T see:
  - B2B vs B2C business model
  - Product vs services/consulting company
  - Active vs defunct/stealth company
  - Competitor exclusion

### Don't Over-Qualify
- Keep criteria loose - we don't want to shrink the list too much
- Qualification is a safety net, not a precision filter
- Better to include marginal prospects than exclude good ones

### Disqualification > Qualification
- Prefer disqualification criteria over qualification criteria
- Only remove prospects with clear, obvious deal-breakers
- Err on the side of qualifying when data is ambiguous

### Data Availability
The qualification AI has access to:
- LinkedIn profile (title, company, location, previous roles, education, skills, summary)
- Company data (industry, headcount, headquarters, description)
- Website metadata (page title and meta description ONLY)

NOT available:
- Full website content, product pages, tech stack
- News articles, press releases, funding data
- Social media activity, events, or any time-sensitive signal

## Common Disqualification Criteria

A menu to pick from, not a checklist to include in full. Always include competitor exclusion; add one or two others only if they're a real risk for this audience. Keep each line short.

- **Competitor** (always): "Company provides [client's service/product category]"
- **Business model**: B2C when targeting B2B; agency/consultancy when targeting product companies
- **Company stage**: pre-revenue/stealth, or acquired/merged
- **Role mismatch**: contractors/freelancers or board/advisors when targeting operators

Remember: none of these should restate a search-filter dimension.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall