marketing-shibata50/claude-code-ultimate-guide
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

docs: integrate Boris Cherny / Lenny's Newsletter insights (4/5)

Browse source 
3 contenus inédits issus de l'interview Boris Cherny (Head of Claude Code)
sur Lenny's Newsletter (19 fev 2026), vérifiés Perplexity :

- Ratio "80% Plan Mode" (empirique, inventeur du produit) → §2.3
- "Build for the Model 6 Months Out" (principe stratégique) → §3.2 CLAUDE.md
- 3 principes d'équipe : underfund / unlimited tokens / go faster → §3.5

+ Fichier d'évaluation (4/5, challenge technical-writer, fact-check)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Florian BRUNIAUX 2026-02-25 18:53:30 +01:00
parent 5e893f3ccd
commit 9dc51b7033
3 changed files with 269 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

34
guide/ultimate-guide.md
View file

@ -2347,6 +2347,9 @@ You: Let's plan this feature before implementing
| Fixing a typo | ❌ No |
| Quick edit to known file | ❌ No |
> **Recommended frequency**: Boris Cherny (Head of Claude Code at Anthropic) starts approximately **80% of tasks in Plan Mode** — letting Claude plan before writing a single line of code. Once the plan is approved, execution is almost always correct on the first try.
> — *Lenny's Newsletter, February 19, 2026*
### Exiting Plan Mode
```
@ -4279,6 +4282,18 @@ Month 3: 50 rules → 50 mistakes prevented + faster onboarding
**Anti-pattern**: Preemptively documenting everything. Instead, treat CLAUDE.md as a **living document** that grows through actual mistakes caught during development.
### Build for the Model 6 Months Out
> **"Don't design your workflows around the limitations of today's model. Build for where the technology will be in six months."**
> — Boris Cherny, Head of Claude Code, Lenny's Newsletter (February 19, 2026)
The corollary: every investment you make today in CLAUDE.md, skills, hooks, and workflows compounds *harder* as the models improve. If you optimize purely for current limitations, you'll be constantly rewriting your setup. If you build for a slightly more capable model, your workflows will run automatically when the next version drops.
**Practical implications**:
- Write CLAUDE.md rules as if Claude will understand nuance better — don't over-specify constraints that will be unnecessary with the next model
- Build agents for goals, not for step-by-step procedures (models get better at navigation, not just execution)
- Invest in your prompt patterns and slash commands now — they age well
### Continuous Context Update
Beyond reactive error capture, **proactively document discoveries** during development sessions. Every insight Claude surfaces about your codebase is a potential CLAUDE.md entry.
@ -5474,6 +5489,25 @@ When multiple developers use Claude Code on the same codebase, hidden AI generat
> **Anti-pattern**: Skipping disclosure to move faster. The hidden cost is reviewers approving code nobody understands, compounding over months into sections of the codebase that are opaque to the whole team.
### Boris Cherny's 3 Principles for AI Teams
> These are the principles Boris Cherny (Head of Claude Code at Anthropic) shares with every new team member.
> — *Lenny's Newsletter, February 19, 2026*
**1. Underfund projects on purpose**
Having one great engineer on a big problem — instead of a full team — forces deep AI utilization. The constraint accelerates shipping, not slows it. The bottleneck shifts from headcount to quality of prompts and workflows.
**2. Give engineers unlimited tokens first**
Don't optimize token costs early. Give engineers the freedom to experiment maximally. Crazy, innovative patterns only emerge when nobody is watching the meter. Optimize costs *after* a successful idea has proven its value and needs to scale.
**3. Encourage people to go faster**
The default instinct with AI tools is caution — reviewing every output, second-guessing every suggestion. The better instinct: ship, validate, iterate. Claude Code is designed for high-velocity cycles, not careful deliberation.
> **When to apply**: Teams of 2+ using Claude Code professionally. Solo developers should focus on the first two principles (underfund = treat yourself as a one-person team with AI leverage; unlimited tokens = don't self-censor your experiments).
---
# 4. Agents