# Cofounder Matching > Rubric-based co-founder matching methodology. Scores candidate co-founders against a founder's startup profile using a 6-axis framework (domain fit, skill complementarity, values alignment, equity expectations, time commitment, track record). Produces a ranked shortlist with gap analysis + interview questions per candidate. Works standalone (manual candidate input) — can consume DojoOS candidate pool via dojoos-api-consumer agent when available. Use when the user asks "co-founder matching", "find co-founder", "evaluate candidate", "co-founder fit", "co-founder scoring", "/cofounder-matching". NOT legal or HR advice. - Skill: `dojocodinglabs/cofounder-matching` (Agent Skill) - Install (CLI): `npx skillmds@latest add dojocodinglabs/cofounder-matching` - Raw SKILL.md: https://api.skillmd.com/api/skills/dojocodinglabs/cofounder-matching/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - Author: dojocodinglabs (https://skillmd.com/u/dojocodinglabs) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/dojocodinglabs/cofounder-matching --- # Co-Founder Matching Evalúa candidatos a **co-founder** contra un startup profile usando una **rubric de 6 ejes**. Produce un ranked shortlist + gap analysis + interview questions customizadas por candidato. ## ⚠️ Disclaimer Co-founder matching es **decisión high-stakes**: - Co-founder equity grants son típicamente 20-50% — irreversible sin término (o clawback tras vesting) - Co-founder disputes son la razón #1 de failure en early stage startups - Este skill genera **marco de evaluación estructurado**, NO sustituye: - Working session trial (4-8 semanas de trabajo pre-grant) - Legal review del Founder Agreement - Background / reference checks profesionales - Red flag conversations (visión a 10 años, sacrifice tolerance, exit expectations) ## Regla de idioma Español. ## Directorio de salida ``` ./launchpad/{startup-slug}/cofounder-matching/ ├── rubric.md # Rubric personalizada para esta startup ├── candidate-[name]/ │ ├── scorecard.md # 6-axis scorecard con evidencia │ ├── gap-analysis.md # Qué falta vs ideal candidate │ └── interview-questions.md # Preguntas customizadas para este candidato └── shortlist.md # Ranked list + recommendation ``` --- ## La rubric — 6 ejes de evaluación ### 1. Domain fit ¿El candidato entiende el mercado objetivo? **Score 1-5**: - 5: Operador/builder en el vertical con >5 años + network relevante - 4: Usuario profundo del problema + research extensa del mercado - 3: Familiar con el mercado pero sin operator experience - 2: Conoce el mercado desde afuera (consumidor casual, articles) - 1: Sin conocimiento relevante del dominio ### 2. Skill complementarity ¿Las skills del candidato llenan gaps del founder existente? Mapear a los 3 roles archetype de Y Combinator: - **Hacker** (builds the product) - **Hustler** (sells and markets) - **Designer** (designs the experience) **Score 1-5**: - 5: Cubre completamente un gap crítico + es world-class en eso - 4: Cubre un gap + tiene competencia sólida - 3: Cubre parcialmente un gap existente - 2: Duplica skills existentes (overlapping, not complementary) - 1: No aporta skills relevantes al stage actual ### 3. Values alignment ¿Comparten los valores clave para esta venture? 5 valores standard a evaluar: - **Ambition** (growth-oriented vs. lifestyle) - **Risk tolerance** (VC path vs. bootstrap) - **Time horizon** (10-year commitment vs. 3-year exit) - **Ethics** (how they've handled moral edge cases pre) - **Impact** (profit alone vs. impact + profit) **Score 1-5**: - 5: Explícitamente alineado en los 5 valores con evidencia - 4: Alineado en 4/5, el 5° es aceptable divergence - 3: Alineado en 3/5, 2 divergencias require explícit conversation - 2: Divergente en ≥3 valores — red flag - 1: Misalignment fundamental que predicto breakdown ### 4. Equity expectations ¿El candidato espera un equity split compatible con el stage + contribution? **Score 1-5**: - 5: Expectativa razonable (10-50% dependiendo stage) + flexible + firma vesting 4yr/1yr cliff sin fricción - 4: Razonable pero negotiations on terms (acceleration, cliff variation) - 3: Expectativa alta pero negociable con clarification - 2: Expectativa inflada (e.g., 50% joining at Seed post-MVP) - 1: Unreasonable — demand sin justificación o sin aceptar vesting ### 5. Time commitment ¿Fulltime es realista para este candidato? **Score 1-5**: - 5: Available full-time desde día 1, financial runway personal ≥12 meses - 4: Full-time en ≤30 días + financial runway ≥6 meses - 3: 80%+ committed con day-job por ≤3 meses transition - 2: Part-time indefinido (>50%) — risky pero manageable para stage - 1: Part-time <30% o indefinido — usualmente red flag para founder ### 6. Track record + execution evidence ¿Qué ha shipped / construido / logrado el candidato antes? **Score 1-5**: - 5: Exit previo o role senior en exitosa startup + refs strong - 4: Built/shipped product con traction cuantificable + refs positivos - 3: Builds independientes con evidence (GitHub, Dribbble, case studies) - 2: Training/certification sin ship evidence - 1: Puro CV sin evidence de ship capability --- ## Weighted scoring Los ejes NO pesan igual — el peso depende del stage de la startup: | Eje | Ideation | MVP | Traction | Funded+ | |---|---|---|---|---| | Domain fit | 20% | 15% | 25% | 30% | | Skill complementarity | 30% | 35% | 25% | 20% | | Values alignment | 25% | 20% | 15% | 15% | | Equity expectations | 10% | 10% | 15% | 15% | | Time commitment | 10% | 15% | 10% | 10% | | Track record | 5% | 5% | 10% | 10% | **Rationale**: early stage prioriza skill complementarity + values (team gelling); later stage prioriza domain fit + proven execution. --- ## Flujo del skill ### Paso 1 — Load startup context **CM-1**: "Vamos a evaluar co-founder candidates contra tu startup. Primero necesito el context: 1. ¿Tenés `startup-profile.md` generado por `startup-intake`? Si sí, lo leo. 2. ¿Cuál es el stage actual? (Ideation / Formation / MVP / Traction / Funded / Scaling) 3. ¿Qué gaps de skills tenés explícitos? (CEO + falta CTO, o tenés CEO+CTO y te falta CMO, etc.) 4. ¿Qué equity range estás considerando ofrecer? (típico 20-50% dependiendo contribution pre-existente)" ### Paso 2 — Generate rubric personalizada **CM-2**: Generar `rubric.md` con weights ajustados al stage, y hacerlo visible al usuario para que vea los criterios antes de evaluar candidates. ### Paso 3 — Collect candidates **CM-3**: "¿Cuántos candidates querés evaluar? Para cada candidate, necesito: - Nombre + contacto (LinkedIn URL ideal) - Cómo los conociste - Domain expertise claim (self-described + evidencia) - Skills claim - Time commitment expectation - Equity expectation (o rango) - Track record: 2-3 highlights verificables Si tenés DojoOS API habilitada (via `dojoos-api-consumer` agent), puedo pull candidates from your co-founder matching queue directly. Si no, proceedemos con manual input." ### Paso 4 — Score cada candidate **CM-4**: Para cada candidate, generar `scorecard.md` con: - Score por eje (1-5 con evidencia citada) - Weighted total score - Relative ranking vs otros candidates - Key strengths - Key concerns - **Kill-switch flags**: criterios que si fallan, desqualify regardless de total score (ej. values misalignment severo, equity expectation outrageous, concurrent commitment a competitor venture) ### Paso 5 — Gap analysis **CM-5**: Generar `gap-analysis.md` por candidate: - Qué axis están under-scored - Can gaps be closed pre-grant? (training, trial period, advisor bridge) - Can gaps be tolerable post-grant? (rare — usually breaks later) - Compare candidate profile to "ideal candidate" for this stage ### Paso 6 — Interview questions **CM-6**: Generar `interview-questions.md` customizado por candidate: - 5-8 preguntas targeted a sus weakest axes - 2-3 preguntas de values exploration (hipotéticos concretos: "Si el lead VC te pidiera firing de un early employee antes de Series A, ¿cómo responderías?") - 1-2 red flag probes (ej. "Contame de un conflict con un co-founder o partner pasado y cómo se resolvió") - Reference check questions para terceros (ex-managers, ex-co-founders) ### Paso 7 — Shortlist + recommendation **CM-7**: Generar `shortlist.md`: - Ranked list por weighted score - Top recommendation con rationale - Runner-up con rationale - "Avoid" list con explicit red flags - Next steps recommended: - **Trial period before grant** (4-8 weeks paid engagement) - **Reference checks** (3+ independent) - **Founder Agreement + Vesting** (via `founder-documents` skill) - **Shared equity milestones** (e.g., grant 5% immediately, rest after 3-month trial validated) --- ## Output template — scorecard.md ```markdown # Candidate Scorecard — [Candidate Name] **Startup**: [Startup Name] **Evaluator**: [Founder Name] **Date**: YYYY-MM-DD **Stage weighting**: [Ideation / MVP / Traction / Funded] --- ## Overview - **Contact**: [LinkedIn / email] - **How we met**: [context] - **Time commitment claim**: [full-time / part-time X%] - **Equity expectation**: [X% range or specific] --- ## Score by axis (1-5 with evidence) ### 1. Domain fit — Score: X/5 (weight XX%) **Evidence**: - [Specific observation, quote, or claim] - [Another evidence] **Analysis**: [why this score, not higher/lower] ### 2. Skill complementarity — Score: X/5 (weight XX%) [same format] ### 3. Values alignment — Score: X/5 (weight XX%) [Evaluate each of the 5 standard values with specific observations] ### 4. Equity expectations — Score: X/5 (weight XX%) ### 5. Time commitment — Score: X/5 (weight XX%) ### 6. Track record — Score: X/5 (weight XX%) --- ## Weighted total | Axis | Score (1-5) | Weight | Weighted | |---|---|---|---| | Domain fit | X | XX% | X.XX | | Skill complementarity | X | XX% | X.XX | | Values alignment | X | XX% | X.XX | | Equity expectations | X | XX% | X.XX | | Time commitment | X | XX% | X.XX | | Track record | X | XX% | X.XX | | **TOTAL** | — | **100%** | **X.XX / 5** | --- ## Kill-switch flags - [ ] Values severe misalignment? Y/N - [ ] Equity demand unreasonable / no vesting? Y/N - [ ] Concurrent commitment to competitor? Y/N - [ ] Failed reference check? Y/N - [ ] IP conflict / non-compete issue? Y/N If ANY kill-switch → **DISQUALIFY regardless of score**. --- ## Summary **Strengths**: - [Top 3] **Concerns**: - [Top 3] **Recommendation**: [Progress to interview / Trial period / Pass] **Next step**: [specific action with timeline] ``` --- ## Integración con DojoOS (via dojoos-api-consumer agent) **Disponible desde v0.5.0** — el agent `dojoos-api-consumer` (ver `agents/dojoos-api-consumer.md`) es invocable desde este skill en el Paso 3 (collect candidates) pidiendo la operación `list_candidate_pool`. Hoy esa operación retorna `SPEC_GAP` (el endpoint no está en la OpenAPI spec todavía) y el skill procede con manual candidate input — como siempre. Cuando @william + @garbanzo expongan el endpoint en DojoOS, el agente comenzará a retornar `LIVE_DATA` automáticamente sin cambios en este skill: candidate pool + Dojo Score embebido se inyectan al rubric 6-axis. Mismo patrón para push de scorecards de vuelta via `sync_candidate_scorecard` (también `SPEC_GAP` hoy). ## Integración con otras skills - **`startup-intake`**: source de `startup-profile.md` para stage + gaps context - **`cap-table-builder`**: post-matching, el co-founder grant se ejecuta vía este skill - **`founder-documents`**: Founder Stock Purchase Agreement + IP Assignment + Vesting Exhibit - **`feature-to-spike`**: si durante matching se descubre pattern útil para DojoOS (ej. "weighted scoring ajustado por stage es mejor que pesos fijos"), generar SPIKE ## Principios clave - **Kill-switches override score**: nunca avanzar si hay values misalignment fundamental - **Trial period obligatorio antes de grant completo**: 4-8 semanas paid + milestone-based equity release - **Reference checks siempre**: mínimo 3 ex-colleagues o ex-partners (NO solo los que el candidate eligió) - **Weighted scoring ajustado por stage**: pesos no son fijos, dependen del momento de la venture - **Documentación como audit trail**: cada scorecard referencable ante future founder disputes ## Anti-patterns - Grant de equity sin vesting porque "confiamos" → red flag VC + future dispute risk - Skipping reference checks porque the candidate es amigo de un amigo - Over-weighting de domain fit en Ideation stage (se aprende; skill complementarity pesa más temprano) - Ignorar kill-switches "porque el candidate es excepcional en otros ejes" - Una sola session de evaluación (no trial period) ## Recursos - **Y Combinator Co-Founder Matching** ([ycombinator.com/cofounder-matching](https://www.ycombinator.com/cofounder-matching)) - **Paul Graham — "The 18 Mistakes That Kill Startups"** — #1 is single founder - **[First Round Review — Co-Founder Guide](https://review.firstround.com/)** - **"The Founder's Dilemmas"** (Noam Wasserman, Princeton 2012) — research sobre founder splits - **Reference check framework**: 3×3 (3 refs, 3 questions each: values, skills, conflict resolution)