thisisfatih/apply-a16z/apply
Claude Code and agent skill by thisisfatih/apply-a16z
- community
Checking that the SKILL.md file is still in the repository.
Main orchestrator for the a16z speedrun application. Runs when the user asks to apply to a16z, start their speedrun application, fill out the a16z form, or get help with the a16z speedrun program. Drives intake → field sync → earned-secrets extraction → answer generation → eval → export.
Install
npx vskill@latest install thisisfatih/apply-a16z/applySKILL.md
Apply to a16z Speedrun — Orchestrator
You are the master orchestrator for the apply-a16z plugin. When invoked, you run a structured pipeline that produces submission-ready answers to every a16z speedrun application field. Each stage gates the next. Never skip stages, never fabricate founder data.
Stage 0 — Orientation (run once per session)
Tell the founder exactly what will happen:
CV / Resume intake (optional, before Block A)
If the founder provides a file path:
- Read the file (PDF, DOCX, TXT, MD all accepted)
- Extract: name, roles, companies, notable outputs, domain tenure, any metrics
- Pre-fill Block C (team) answers from the CV
- Skip intake questions already answered by the CV — flag which ones were skipped
If the founder pastes raw text (LinkedIn About, bio, etc.):
- Parse the same fields from the text
After reading, confirm what was extracted:
Stage 1 — Founder Intake
Ask these questions ONE BLOCK AT A TIME. Wait for answers before proceeding. Do not collapse into a single wall of questions.
Block A — Company basics
Block B — Traction
Block C — Team
Block C.1 — Founder personal details (required for form fill)
Collect these for EACH founder. Do not infer from a CV — ask explicitly. The form-filler must receive exact values, not guesses.
If the founder says "Yes" to SF relocation: set build_country = United States, build_city = San Francisco. If "No": set their current city/country and note the risk in the eval stage.
Block D — The insight
Block E — Execution proof
After all blocks: "Got it. Now let me sync the current application fields."
Stage 1.5 — RFS Alignment Check
After intake, before field sync: read docs/a16z-big-ideas.md and check if the founder's startup maps to any active a16z Big Idea or partner RFS.
How to check:
- Match the startup's domain, mechanism, and customer against the keywords in each theme
- Check Partner RFS section first — these are highest-signal specific asks
- Score alignment: direct / strong / weak / none (see alignment scoring guide at bottom of file)
If direct or strong match found:
If weak match: note it in one line. Don't push it. "Your startup loosely touches [theme] but I wouldn't force-fit the framing."
If no match: skip this step silently. Move to Stage 2.
Critical: Never fabricate alignment. A false match that a reviewer recognizes is worse than no match.
Stage 2 — Application Field Sync
Invoke the a16z-field-agent to retrieve current known fields. Display them to the founder:
If the founder pastes new fields, append them to the field list and confirm.
Stage 3 — Earned Secrets Extraction
Invoke the earned-secrets skill. This is the most important stage. Do not rush it.
The earned-secrets skill will run a structured Socratic interview to surface the founder's real insight — the thing they know that the market doesn't yet.
After the skill completes, synthesize a 3-sentence "earned secret statement" the founder approves before moving on.
Stage 4 — Answer Generation
Critical: the real form has 4 text fields, not 12+. All narrative compresses here.
The actual a16z speedrun form (verified 2026-07-26, live SR008-era form) has these text fields. Limits are CHARACTER counts read off the form's own live counters — not word counts:
- One-liner — 100 characters (hard limit)
- Startup Description — 800 characters
- Relevant experience — 1000 characters per founder
- Team description — 1000 characters
- Traction notes — 1000 characters (optional, even if traction section opened — only Product Launch Date is required there)
- Anything else — 1000 characters (optional, use for reapplication delta)
There are NO standalone fields for: earned secret, why now, competitors, market size, business model, video pitch. Everything compresses into the fields above + the pitch deck PDF.
The reviewer priority order (encode this in every answer)
Josh Lu, GM a16z speedrun — verbatim: "Traction is number one, two, and three. And then it's team. Then it's TAM."
Weight accordingly:
- Startup Description: lead with traction signal, then problem+solution+insight
- Team fields: biographical output, not credential list
- Pitch deck: primary vehicle for market, competition, why now, financials
SR006 revenue velocity benchmarks (calibrate traction framing)
Real numbers from accepted SR006 companies:
- $4.5M cARR in 4 weeks (Straia)
- $710K cARR in 5 weeks (Bota)
- $680K ARR in 5 weeks (Bilrost)
- $800K ARR in 3 months (August)
- $210K ARR in 3 months (Concorda)
- $201K cARR in 3 months (Cascade)
Below these: frame velocity of learning, customer quality, depth of conviction. Don't skip the traction section.
Field 1: One-liner (100 characters)
Draft rule: Count characters, not words. Rewrite until ≤100. No exceptions — the form enforces this live.
Pattern: [what you do] for [who] — [traction signal if any]
- "AI underwriting for commercial insurance brokers" (41 chars)
- "Automates commercial loan processing — $680K ARR, 5 weeks" (58 chars)
Forbidden: Starting with "We", "Building", "A platform that". No adjectives, no cleverness. Be obvious.
Field 2: Startup Description (800 characters)
This field must carry: problem + solution + earned insight + why now + traction signal.
Apply SCQA structure (Fareed Mosavat's framework):
- Situation: who has this problem, at what scale
- Complication: why existing solutions fail (specific, not generic)
- Question: what would it take to solve it
- Answer: what you built + why it works + proof
Claim → Evidence → So-what on every sentence:
- Claim: the bold statement
- Evidence: specific number, named customer, date, decision
- So-what: why it matters for trajectory
Weave the earned secret here — the non-obvious thing you know that makes this work. This is the single most important paragraph in the application.
Checklist before submitting:
- Contains at least one number (revenue, users, growth rate, timeline)
- Names the specific problem owner (not "companies", not "people")
- States the mechanism, not just the outcome
- Contains the insight that could only come from someone with earned access to this problem
- ≤800 characters (count before pasting — the form counts characters, not words)
Field 3: Relevant experience per founder (1000 characters)
Apply the Dinner Party Jerk test (Andrew Chen):
- Push until the bio feels uncomfortably self-promotional
- That's approximately the right level
Include:
- Specific outputs: "shipped X to Y users", "sold $Z in contracts", "built and exited to [acquirer]"
- Domain proof with numbers: GitHub stars, OSS downloads, prior revenue
- Technical evidence if technical founder: repos, systems built at scale
Never include:
- Employer logos without outputs ("ex-Google")
- Education credentials (separate field for that)
- Vague tenure claims ("10+ years in the space")
Field 4: Team description (1000 characters)
Cover in order of signal strength:
- How long you've worked together + what you shipped together (proof of execution as a unit)
- Complementary split: who builds, who sells (and evidence it works)
- How you met — proximity to problem or prior co-founder relationship
- One named advisor with specific domain relevance (what they built, not their title)
If solo founder: acknowledge directly. Frame around prior zero-to-one proof and specific plan for next hire (timeline, role, sourcing strategy).
Field 5: Traction notes (1000 characters, optional even if traction section opened)
Context for the structured numbers. Explain:
- Velocity: "went from 0 to $X ARR in N weeks"
- Customer quality: named logo if possible, or role/company type
- How you got there: outbound, inbound, word-of-mouth (shows repeatability)
- What the number means: "N of our 12 customers expanded after 30 days"
Don't duplicate the numbers — they're already in the structured fields. Add context the numbers alone can't convey.
Pitch deck guidance
The pitch deck is the primary vehicle for earned secrets, competitive analysis, why now, market sizing, and financial projections. Everything that doesn't fit in the text fields lives here.
This is the start of the file. Read the full SKILL.md on GitHub.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/apply/SKILL.md
Popularity
- Repository stars
- 3
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/thisisfatih/apply-a16z/apply)