andreessen
by alirezarezvaniandreessen is a market-first Strategic Planning skill for pressure-testing ventures, product ideas, features, PMF signals, career bets, and daily priorities. It uses an anti-sycophancy operating prompt, verdict labels, reference guides, examples, and stdlib Python scripts for market scoring, PMF checks, and 3x5 Anti-Todo execution.
This skill scores 84/100, which means it is a solid listing candidate for directory users who want a reusable market-first venture, product, or productivity pressure-test workflow. The repository gives enough evidence to understand when to invoke it, what outputs to expect, and how scripts/examples support execution, though installation guidance and fit boundaries could be clearer.
- Very clear triggers in frontmatter, including “andreessen,” “pmarca mode,” “should I build this,” “pmf check,” and market-first pressure-test requests.
- Provides concrete workflows and artifacts: market verdicts, PMF checks, forcing-question worksheet, 3x5 card, Anti-Todo routine, and worked examples.
- Includes three stdlib Python scripts for deterministic scoring/card routines, with README usage examples and JSON output support.
- No install command is present in the skill package, so users must rely on the broader repository/plugin instructions to adopt it.
- The value depends heavily on wanting a blunt Marc Andreessen-style persona; it is explicitly “not for polite brainstorming” and may be a poor fit for collaborative or nuanced exploration.
Overview of andreessen skill
What andreessen is for
andreessen is a Claude skill for blunt, market-first Strategic Planning. It pushes an agent to evaluate ventures, product ideas, features, career bets, and daily priorities through a Marc Andreessen-inspired lens: the market matters more than team or product, product/market fit is the decisive milestone, and unclear demand should be treated as a warning rather than an invitation to brainstorm.
Best-fit users and decisions
The andreessen skill is strongest for founders, product leaders, investors, builders, and operators who want a hard verdict before committing time, capital, or team attention. Use it when asking: “Should we build this?”, “Is this a real market?”, “Are we before or after PMF?”, “Should we kill or pivot this feature?”, or “What should I do today if PMF is the only goal?”
What makes it different
Unlike a generic strategy prompt, andreessen has an explicit anti-sycophancy operating style: strongest counterargument first, no premise validation, explicit confidence levels, and no softening when the evidence is weak. The repository also includes deterministic helper scripts for market scoring, PMF signal scoring, and the 3x5 card plus Anti-Todo productivity routine.
Main adoption tradeoff
This is not a polite ideation skill. It is designed to be adversarial, compressed, and judgment-heavy. That is useful when you need a market-first verdict, but a poor fit if you want broad creative exploration, brand messaging, team morale support, or consensus-building language.
How to Use andreessen skill
andreessen install context
Install from the skill package path in alirezarezvani/claude-skills using your skill manager, for example:
npx skills add alirezarezvani/claude-skills --skill andreessen
The upstream SKILL.md emphasizes portability: the reasoning skill works in Claude Code CLI and Claude.ai, and its Python scripts use the standard library only. There are no external APIs or model calls inside the scripts, which lowers setup friction and makes the scoring tools easy to inspect.
Files to read before first use
Start with SKILL.md to understand routing and expected behavior. Then read references/operating_prompt.md because the voice constraints are central to the skill. For decision substance, inspect references/market_first_canon.md and references/pmf_and_build_canon.md. For examples, open assets/example_market_verdict.md, assets/example_pmf_check.md, and assets/forcing_question_worksheet.md.
If you plan to use the productivity mode, preview assets/blank_3x5_card.md, assets/example_3x5_card.md, and references/personal_productivity_system.md.
Inputs that produce better verdicts
For strong andreessen usage, do not ask “Is this a good idea?” in isolation. Provide the market, buyer, budget, urgency, current alternatives, why now, evidence of pull, retention or usage signals, acquisition mix, and what you are unwilling to change.
Weak prompt:
andreessen, should we build an AI compliance tool?
Better prompt:
andreessen, pressure-test this venture. Market: US medical-device startups preparing FDA submissions. Buyer: VP Regulatory or founder. Budget: currently spent on consultants. Why now: LLMs can draft structured documentation. Evidence: 8 interviews, 2 paid pilots, no self-serve demand yet. Team: two ex-FDA reviewers and one ML engineer. Concern: buyers may distrust AI output. Give a market-first verdict, strongest counterargument first, confidence level, and the one next move.
Practical workflow with scripts
Use the skill conversationally for the strategic judgment, then use scripts when you want repeatable scoring. Example repository commands include:
python scripts/market_first_evaluator.py --size 8 --growth 7 --timing 9 --pull 8 --team 6 --product 5
python scripts/pmf_signal_scorer.py --ellis-pct 45 --retention 8 --organic 7 --demand 8 --frequency 7
For daily execution, use:
python scripts/anti_todo_card.py --new --must-do "Call 5 churned users" "Ship retention dashboard" "Cut onboarding to 3 steps"
The best workflow is: run the market or PMF verdict, identify the dominant strategic variable, then write a 3x5 card where every must-do moves that variable.
andreessen skill FAQ
Is andreessen for Strategic Planning or productivity?
Both, but Strategic Planning comes first. The productivity system is downstream of the verdict. The skill asks what market, PMF, or leverage problem matters most, then turns that into a capped 3-5 item daily card. If your daily tasks are disconnected from the strategic diagnosis, you are not using andreessen as intended.
How is this better than an ordinary prompt?
A normal prompt can ask for startup advice, but it often mirrors your framing. The andreessen skill has a fixed operating posture: counterargument first, market dominance, PMF gates, confidence levels, and concrete verdict labels such as BUILD-POUR-FUEL, MARKET-FIRST-DERISK, and KILL-OR-REPICK-MARKET. That consistency is the value.
Can beginners use the andreessen skill?
Yes, if they can provide concrete evidence. Beginners should start with assets/forcing_question_worksheet.md because it exposes missing information quickly. If you cannot name the buyer, budget, urgency, and PMF signal, the skill will likely return a harsh but useful diagnosis: you are describing a product, not a market.
When should I not use andreessen?
Do not use andreessen when you need user empathy research synthesis, design exploration, fundraising narrative polish, legal review, or a balanced facilitation voice. It also should not be treated as an oracle. The deterministic scripts are scoring aids, not proof that a market exists.
How to Improve andreessen skill
Improve andreessen inputs before asking
The fastest way to improve andreessen output is to replace adjectives with evidence. “Huge market” is weak; “8,000 target buyers, $40K annual current spend, regulatory deadline in Q3, three buyers asked for procurement docs” is strong. The skill rewards market pull, specific timing, observed behavior, and willingness to change segment, product, or team.
Watch for common failure modes
The main failure mode is treating the skill as a persona for entertaining bluntness instead of a decision system. Another is feeding it vanity metrics: demo praise, waitlists, pitch enthusiasm, NPS without retention, or paid growth without organic pull. For PMF checks, include retention, frequency, organic demand, Sean Ellis survey percentage if available, and whether customers are pulling the product faster than you can supply it.
Iterate after the first verdict
A good first output should create sharper follow-up questions. If the verdict is MARKET-FIRST-DERISK, ask for the riskiest assumption, the cheapest test, the kill criteria, and the next 7-day evidence plan. If the verdict is BEFORE-PMF, ask which segment, use case, or product surface should change first. If the verdict is BUILD-POUR-FUEL, ask what bottleneck to remove before scaling.
Customize without weakening the edge
If adapting the skill, preserve the anti-sycophancy rules and verdict structure. You can add your company’s market definitions, PMF metrics, pricing thresholds, or category-specific constraints, but avoid turning the voice into generic coaching. The value of andreessen is that it refuses to confuse effort, taste, or team quality with market pull.