conon.ai

conon.ai/Tips

Keys Live In .env, Never In Chat

A pasted key sits in chat history under whatever retention your settings dictate. Instead: 'create a .env file for secrets, add it to .gitignore, and verify it can never reach GitHub'. Then point the agent at the file — 'the key lives in .env, take it from there' — so the secret never crosses the chat.

AdvancedTOOLSClaude Code

Why it works

A key pasted into chat becomes chat history — stored under whatever retention your settings dictate, far outliving the moment. The .env pattern separates location from value: secrets live in one file, the file is excluded from version control by name, and the agent gets DIRECTIONS to the key rather than the key itself. One sentence sets the whole thing up, plus a proof.

How to do it

  1. Say: 'create a .env file for this project's secrets and add it to .gitignore.'
  2. Demand the proof, not the promise: 'verify the file can never be committed or pushed.'
  3. Put keys into the file yourself (or paste them there directly — the file, not the chat).
  4. From then on, reference by location: 'the API key is in .env under STRIPE_KEY — use it from there.'
  5. Grep your own history once: any key you EVER pasted into a chat gets rotated today.
  6. New machine or teammate? .env travels by secure channel, never by message.

Copy this

prompt
Create a .env file for this project's secrets, add it to .gitignore, and verify — show me — that it can never be committed or pushed to GitHub.

What to watch for

  • .gitignore protects the repository, not the disk: an agent working in the folder can still read .env. That's the design — it needs the key to use the key — but it means work folders hold work secrets only.
  • Screenshots count as pasting: an image of a terminal with a key in it is the same leak with worse searchability.

Sources

Original: conon.aiDocumentation: code.claude.com

Read next

Scheduled Agents Need An Open Lid
Local scheduled tasks run only while the app is open and the machine is awake — a missed slot is skipped, not postponed. Turn on keep-awake, leave the laptop lid open, and write drift-tolerant instructions: 'if it's already past 5pm, skip today and report that you skipped'. Need lid-closed reliability? That's a cloud task — which can't see your local files.
Claude Code
Typed It Three Times? Make A Skill
The third time you type the same procedure is the signal: say 'create a skill that does this' and the agent writes the skill file itself. Then improve it in a loop — run it, note what's off, have it edit its own skill. House rules: 2nd repeated mistake → CLAUDE.md line; 3rd repeated prompt → skill.
Claude Code
Writer Session, Reviewer Session
Run two parallel sessions on one version-controlled project: one builds, the other — with fresh context — reviews. The docs are explicit about why it works: a fresh session isn't attached to code it didn't write, so it critiques honestly where the author session defends.
Claude Code
Send Noisy Research To A Subagent
Any side-investigation that would flood your session — 'which of these fifty files mentions the old pricing?' — goes to a subagent. It reads everything inside its own separate memory and returns three lines; your session's context stays clean for the actual work.
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 →