agentleFS
samsational-codesamsational-code/business-skills/skill-idea-generator/SKILL.md

skill-idea-generator

Skill0 starsChanged 2 months agoMITView on GitHub

Brainstorm, score, and prioritize new AI skill ideas for any team by turning recurring workflows, bottlenecks, and professional judgment loops into concrete, buildable, reusable skill concepts. Use this skill whenever a manager, operator, analyst, salesperson, marketer, product lead, consultant, clinician, fundraiser, founder, or executive wants to identify new AI workflows worth building for their team. Trigger on phrases like "help me come up with skill ideas," "what AI workflows should we build," "what skills would be useful for our team," "what skills would help our sales / support / finance / research team," "generate AI use cases for my company," "find workflow ideas," "turn this process into a skill," "what should we add to our skills library," or any request to brainstorm, prioritize, or design new reusable AI skills for research, analysis, outreach, communication, monitoring, reporting, operations, or internal productivity. Output is a diagnosis, 8–12 scored skill idea cards, a ranked top 3, one recommended first build, and an optional starter SKILL.md outline. Do NOT use for a structured interview-based audit of a team's workflows that ends in an adoption roadmap — use the ai-workflow-audit skill.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 11 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not generate generic AI ideas.

    line 143
  2. Distinctive:

    Ask only the questions needed.

    line 168
  3. Distinctive:

    If the user says the team spans both archetypes, design ideas that either:

    line 183
  4. Distinctive:

    Call transcript action extractor

    line 262
  5. Distinctive:

    Do not assume AI replaces professional judgment.

    line 295
  6. Distinctive:

    Avoid skills that require unavailable private data, complex integrations or unclear user behavior unless the user explicitly wants an …

    line 322
  7. Distinctive:

    When generating skill ideas, produce:

    line 327
  8. Distinctive:

    Use this format for each idea:

    line 338
  9. Distinctive:

    After presenting all ideas, rank the top 3 using this format:

    line 380
  10. Distinctive:

    Write like a practical AI consultant for professional teams.

    line 503
  11. Distinctive:

    Before delivering the final output, confirm:

    line 524

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill Idea Generator

Generate practical AI skill ideas for any team.

This skill helps professionals turn vague workflow pain into specific, reusable AI skills. The goal is not to brainstorm generic AI use cases. The goal is to identify concrete workflows that are frequent, valuable, repeatable, source-aware and safe enough to build into a skill library.

Why This Skill Exists

Most professionals know where their work feels slow, repetitive or inconsistent, but they do not naturally think in "AI skill" terms.

They say things like:

  • "We waste time preparing for meetings."
  • "Our team summarizes the same kinds of documents over and over."
  • "Our monthly updates to the board are painful to write."
  • "Our CRM notes are a mess."
  • "Every RFP response starts from scratch."
  • "Our first-pass review of vendor proposals is inconsistent."
  • "I want to know where AI can actually help our team."

This skill converts those pain points into well-formed skill ideas.

A good skill idea should have:

  1. A specific user
  2. A recurring workflow
  3. Clear inputs
  4. A useful output
  5. A repeatable process
  6. A quality bar
  7. Known risks and controls
  8. A strong reason to exist

Primary Users

This skill supports two broad archetypes of professional work, plus the many roles that mix them.

Analysis and Decision Roles

Work whose core is gathering, evaluating, and deciding.

Examples:

  • Business, financial, and data analysts
  • Product managers and researchers
  • Consultants and strategy teams
  • Underwriters, credit and risk teams
  • Procurement and vendor-management teams
  • Clinicians and clinical operations leads
  • Engineers and technical leads
  • Auditors, compliance, and legal teams
  • Investors and investment committees

Common workflows:

  • Company, market, and competitor research
  • Vendor, partner, or counterparty evaluation
  • Proposal and RFP-response review
  • Decision memo preparation
  • Performance and variance review
  • Risk review
  • Public-document and filings review
  • Meeting preparation
  • Diligence and interview question generation

Relationship and Communication Roles

Work whose core is understanding people and persuading or informing them.

Examples:

  • Sales and business development
  • Account management and customer success
  • Marketing and communications
  • Fundraising and development (nonprofit or venture)
  • Investor relations
  • Recruiting and talent
  • Partnerships and alliances
  • Client service and relationship management

Common workflows:

  • Prospect and stakeholder research
  • Meeting prep for a key account, donor, or candidate
  • Message personalization
  • Positioning and pitch refinement
  • Objection handling
  • Post-meeting follow-up
  • CRM cleanup
  • Periodic stakeholder updates
  • Questionnaire and RFP responses
  • Thought leadership

Both Archetypes

Some skills serve both kinds of work — and most leadership, operations, and founder roles live in both.

Examples:

  • Strategy or product explainer
  • Industry commentary builder
  • Periodic update generator
  • Meeting transcript follow-up
  • Professional profile brief
  • Business one-pager builder
  • FAQ generator
  • Stakeholder question answerer

Core Principle

Do not generate generic AI ideas.

Bad ideas:

  • "Summarize documents"
  • "Help with research"
  • "Write better emails"
  • "Analyze companies"
  • "Use AI for sales"

Good ideas:

  • "Turn a raw all-hands or earnings-call transcript into a structured memo with key metrics, period-over-period changes, the full Q&A, and where leadership was evasive."
  • "Create a first-pass diligence note on a new vendor that separates known facts, unknowns, and questions to ask in the next meeting."
  • "Turn raw monthly inputs into a polished board or investor update draft with missing-data placeholders."
  • "Generate a pre-meeting profile brief on a prospect, candidate, or donor using public information and user-provided notes."
  • "Convert messy CRM notes into a clean relationship status, next steps, and follow-up drafts."
  • "Draft answers to an incoming security questionnaire from the approved answer library, flagging every question with no approved answer."

The skill idea should sound like something that belongs in a professional workflow library.

First Step: Determine User Context

Begin by identifying the user's role and workflow setting.

Ask only the questions needed. If the user already provides enough context, proceed with reasonable assumptions and state them.

Role Router

Ask:

  1. Is your work primarily analysis and decisions, relationships and communication, operations and delivery, or a mix?
  2. What type of organization is this for? Examples: software company, agency, professional services firm, hospital or clinic, manufacturer, nonprofit, government agency, startup, consultancy, law firm, school or university, investment firm.
  3. What function or team? Examples: sales, marketing, product, engineering, operations, finance, HR, legal, customer success, research, clinical, fundraising, leadership.
  4. What recurring workflow feels slow, annoying, inconsistent or too dependent on one person's judgment?
  5. What materials do you usually work with? Examples: reports, transcripts, decks, questionnaires and RFPs, CRM notes, support tickets, contracts, spreadsheets, survey data, websites, emails, filings, research reports.
  6. What would a useful output look like? Examples: memo, checklist, email, table, scorecard, brief, question list, summary, one-pager, redline.
  7. Is the goal better judgment, faster execution, better communication, better monitoring, more consistency, or internal productivity?

If the user says the team spans both archetypes, design ideas that either:

  • work for both groups, or
  • have separate analysis and communication variants.

Skill Categories

Classify each idea into one category.

1. Meeting and Inbox Prep

Skills that help busy professionals quickly understand what matters before a meeting, call or workday.

Examples:

  • Inbox triage
  • Executive briefing
  • Professional profile brief
  • Meeting transcript follow-up
  • Relationship status brief
  • Pre-call prep note

2. Research and Analysis

Skills that help teams gather, structure and analyze decision-relevant information.

Examples:

  • Company first pass
  • Private company tear sheet
  • Counterparty diligence note
  • Presentation transcript extractor
  • Industry primer
  • Competitive landscape map
  • Plan or thesis stress tester
  • Decision-meeting question generator
  • Red flag detector

3. Outreach and Stakeholder Communication

Skills that help teams communicate with customers, prospects, partners, investors, donors, members, or candidates.

Examples:

  • Stakeholder update generator
  • Business one-pager
  • Intro email generator
  • Key-account meeting prep brief
  • Objection response builder
  • Prospect personalization memo
  • Questionnaire and RFP response drafter

4. Monitoring and Tracking

Skills that help teams track companies, vendors, customers, projects, or open questions over time.

Examples:

  • Project or account update tracker
  • Vendor monitoring brief
  • KPI variance explainer
  • Relationship tracker
  • Follow-up gap detector
  • News change detector
  • Document version diff

5. Internal Productivity

Skills that reduce recurring administrative or knowledge-work friction.

Examples:

  • Project folder organizer
  • CRM note cleaner
  • Weekly priorities generator
  • Call transcript action extractor
  • Document comparison assistant
  • Knowledge base organizer
  • Audit request binder

6. Thought Leadership and Marketing

Skills that help professionals turn expertise into credible public or client-facing content.

Examples:

  • LinkedIn post generator
  • Industry commentary drafter
  • Newsletter idea generator
  • Client education brief
  • White paper outline builder
  • Research note repurposer

Ideation Method

For each workflow, work through the following sequence:

  1. Identify the repeated task.
  2. Identify the expensive human judgment involved.
  3. Identify the source materials.
  4. Identify the desired output.
  5. Identify the current bottleneck.
  6. Identify the cost of doing the work poorly.
  7. Identify the compliance, privacy, hallucination or reputational risks.
  8. Identify the smallest useful version of the skill.
  9. Decide whether the idea belongs in the library.

Do not assume AI replaces professional judgment. Frame skills as leverage for expert professionals.

Scoring Rubric

Score each skill idea from 1 to 5.

DimensionMeaning
FrequencyHow often the workflow happens
PainHow frustrating, slow or costly it is
Judgment leverageWhether it improves high-value thinking
Data availabilityWhether the required inputs are easy to access
RepeatabilityWhether the workflow follows a pattern
Risk manageabilityWhether compliance, privacy and hallucination risks can be controlled
Demo valueWhether the skill is impressive and easy to show

Prioritize skills with:

  • High frequency
  • High pain
  • High repeatability
  • Clear inputs
  • Useful outputs
  • Manageable risk
  • Strong demo value

Avoid skills that require unavailable private data, complex integrations or unclear user behavior unless the user explicitly wants an advanced build.

Output Format

When generating skill ideas, produce:

  1. A brief diagnosis of the user's context
  2. 8 to 12 skill ideas
  3. A ranked top 3
  4. One recommended first skill to build
  5. A short explanation of why that skill is the best starting point
  6. Optional: a starter SKILL.md outline for the top idea

Skill Idea Card Format

Use this format for each idea:

Skill Idea: [Name]

Primary user:
Analysis and decision role / Relationship and communication role / Both

Category:
Meeting and Inbox Prep / Research and Analysis / Outreach and Stakeholder Communication / Monitoring and Tracking / Internal Productivity / Thought Leadership and Marketing

Workflow pain:
[Describe the recurring problem.]

Trigger:
[When the user would invoke this skill.]

Inputs:
[List the materials the user would provide.]

Output:
[Describe the deliverable.]

Workflow:

  1. [Step one]
  2. [Step two]
  3. [Step three]
  4. [Step four]

Quality bar:
[What a good output must include.]

Risks / controls:
[Where the model could hallucinate, overstate, misclassify, leak confidential data, or create compliance risk. Include practical controls.]

Build priority:
High / Medium / Low

Why this is valuable:
[One sentence explaining the business value.]

Ranked Top 3 Format

After presenting all ideas, rank the top 3 using this format:

RankSkillPrimary UserWhy It Ranks HighlyMain Risk
1[Skill name][User][Reason][Risk]
2[Skill name][User][Reason][Risk]
3[Skill name][User][Reason][Risk]

Then recommend one first build.

Guidance for Analysis-Focused Ideas

For analysis and decision roles, favor skills that improve:

  • Research speed
  • Evaluation quality
  • Decision-meeting preparation
  • Risk identification
  • Ongoing monitoring
  • Memo quality
  • Quality of internal debate
  • Follow-up discipline

Useful idea patterns:

  • First-pass research
  • Compare this period to prior period
  • Extract metrics
  • Separate claims from evidence
  • Generate questions
  • Identify gaps
  • Stress-test the plan
  • Track changes over time
  • Convert messy materials into decision-ready memos

Example ideas:

  • Vendor Proposal Variant-Perception Finder (what the proposal emphasizes vs. avoids)
  • All-Hands or Earnings-Call Change Detector
  • Key Supplier Meeting Prep Brief
  • Decision Memo Question Generator
  • Project Portfolio Monitoring Brief
  • Strategy or Thesis Stress Tester
  • Red Flag and Missing Data Detector
  • Industry Primer Builder
  • Clinical Protocol Change Summarizer
  • Contract Obligations Extractor

Guidance for Relationship-Focused Ideas

For sales, marketing, fundraising, recruiting, and account roles, favor skills that improve:

  • Prospect and stakeholder understanding
  • Message personalization
  • Meeting preparation
  • Follow-up quality
  • Objection handling
  • Positioning
  • Update clarity
  • CRM discipline
  • Thought leadership

Useful idea patterns:

  • Prepare for this account / donor / candidate
  • Tailor this message
  • Turn notes into follow-up
  • Convert product or program materials into clear positioning
  • Generate likely objections
  • Summarize relationship status
  • Clean CRM records
  • Draft stakeholder communications

Example ideas:

  • Key Account Meeting Prep Brief
  • Prospect Personalization Memo
  • Objection Response Builder
  • Post-Meeting Follow-Up Drafter
  • CRM Note Cleaner
  • Relationship Status Tracker
  • RFP and Questionnaire Response Assistant
  • Positioning Refiner
  • Donor or Member Update Drafter
  • Candidate Outreach Personalizer

Guidance for Both Archetypes

Some of the best library skills work across both kinds of work.

Favor both-archetype skills when the workflow sits between internal analysis and external communication.

Example ideas:

  • Strategy or Product Explainer
  • Industry Commentary Builder
  • Periodic Update Drafter
  • Professional Profile Brief
  • Meeting Transcript Follow-Up
  • Narrative Refiner
  • FAQ Builder
  • Stakeholder Question Answerer

For these, explicitly show how the same skill adapts:

ModeHow the skill behaves
Analysis modeFocuses on evidence, evaluation, risk and decision-making
Communication modeFocuses on positioning, audience relevance, objections and clarity

What This Skill Does Not Do

  • It does not generate generic AI use cases.
  • It does not recommend building every idea.
  • It does not assume access to private databases, CRMs, ticketing systems, EHRs, data rooms or proprietary systems.
  • It does not ignore compliance, confidentiality, privacy or reputational risk.
  • It does not treat analysis work and relationship work as the same job.
  • It does not imply AI replaces professional, managerial, clinical, legal or compliance judgment.

Tone

Write like a practical AI consultant for professional teams.

Default to:

  • Specific
  • Domain-native — use the team's own vocabulary, not generic AI language
  • Commercially aware
  • Concise
  • Skeptical without being cynical
  • Useful rather than exhaustive

Avoid:

  • AI hype
  • Generic productivity language
  • Overly technical jargon
  • Vague transformation claims
  • "This will revolutionize your company" language

Quality Checks Before Finalizing

Before delivering the final output, confirm:

  • Each idea has a clear user
  • Analysis and relationship users are distinguished where the team spans both
  • The inputs are realistic for this organization type
  • The output is concrete
  • The workflow is repeatable
  • The value is obvious
  • Risks and controls are named
  • The top 3 are actually prioritized, not just listed
  • At least one idea could be built immediately as a SKILL.md
  • The ideas feel consistent with a practical skill library for this team's industry

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.