conon.ai

conon.ai/Tips

/init Writes The Project Briefing

Run /init in a new project and the agent studies your folder and writes CLAUDE.md — a briefing it re-reads at the start of every session. Without it, each session re-explores your project from zero, on your token bill.

IntermediateTOOLSClaude Code

Why it works

Every session starts with amnesia — close the app and the agent forgets your project exists. CLAUDE.md is the briefing it re-reads at every session start: what this project is, how you like things done, what to never touch. Without it, each session re-explores your folder from zero, on your token bill; with it, session two starts where session one left off.

How to do it

  1. In any project, type /init in the prompt field.
  2. The agent studies your folder and writes CLAUDE.md — a plain text file you can open anywhere.
  3. Read it. Delete anything it could re-learn by looking at the folder.
  4. Add what it can NOT infer: your rules, your tone, forbidden moves ('never touch pricing.csv').
  5. When it repeats a mistake, ask: 'add a rule to CLAUDE.md so this never happens again.'
  6. Re-run /init after big changes — it proposes updates rather than overwriting your rules.

Copy this

prompt
/init

What to watch for

  • Writing the briefing by hand from scratch wastes the feature — let /init draft it from evidence, then edit.
  • A briefing that restates the obvious ('this folder contains HTML files') is pure token tax; every line should earn its re-reading.

Sources

Original: conon.aiDocumentation: code.claude.com

Read next

Make It Review Its Own Diff
'Done' is a claim, not a fact. Before accepting changes say: 'check these changes before I accept — look for real errors, not formatting'. The agent regularly finds its own bugs, and the closing phrase stops it nitpicking commas instead of logic.
Claude Code
Show A Screenshot, Skip Adjectives
Words like 'modern and minimalist' mean nothing — every ugly site's brief said that. Saw a style you like? Screenshot it into the prompt with 'in this spirit'. A picture transfers taste that three paragraphs of adjectives never will.
Claude Code
Explanatory Style For Month One
Ask for the explanatory output style and the agent narrates why it chose what it chose as it works — the black box becomes a running commentary. When the explanations start to annoy you, that's your graduation signal: switch back to default.
Claude Code
More Effort, Or A Bigger Model?
Bad result? Diagnose before you rage-quit to another tool. Skipped files, no self-check, quit halfway — that's low effort: raise the effort level. Read everything and still confidently wrong — that's the model's ceiling: escalate the model. This one question settles 95% of 'the AI got dumber' complaints.
Claude Code

We build and operate the systems behind tips like this one — agents, automations and internal tools, shipped into production. See what the studio does →