After 28 OKR cycles, this founder shares the Claude Skill he uses to review team OKRs
lukeharries · x · 2026-10-01
lukeharries shares the full Claude Skill he refined over 7 years (28 OKR cycles) to review and improve team OKR plans.
Key rules in the skill:
- OKRs should form a metric tree flowing down from company KPIs, with MECE objectives and covered critical drivers
- Max 4-5 objectives per team (never more than headcount), each person in one objective, max 3-4 KRs per objective, always assessable — no vague "improve X"
- Product eng teams should center on what they want to ship; growth teams need a mix of output metrics and input actions with SMART goals (e.g. launch an AI landing-page system and ship 20 pages by end of November, lifting signup conversion from 3.1% to 4%)
- Think in systems and agentically, building compounding systems; teams should own what they directly control; keep hiring out of OKRs unless it's a critical bottleneck
More from coding & agent
- Engineer pushes back on AI swarm hype: 10K sessions for days is wasteful — gerardsans · 2026-11-16
- "Nuance tends to lead to slop": poteto's talk says design APIs to remove agent decision freedom — mattpocockuk · 2026-11-16
- What Labs Actually Do With $100M Data Deals: Turning Company Workflows Into Agent Eval Sandboxes — StewartalsopIII · 2026-11-16
- Google donates Agent Substrate, gVisor and OpenRL to CNCF; Agent Substrate 1.0 goes GA on GKE — dl_weekly · 2026-11-16
- Plonge: open-source tool safely cleans coding-agent caches, freed 40GB on one Mac — JordanMorgan10 · 2026-11-16
- Teknium: avoid 900k GPT variants — they charge extra beyond 272k context — Teknium · 2026-11-16