conon.ai

conon.ai/Tips

No Connector Listed? Hand It Docs

The connector catalog is a storefront; thousands more MCP servers exist. Paste the server's documentation link and say 'wire this up, figure out the config yourself, and verify it works' — the agent edits its own config files. Only install servers from authors you trust: a malicious server is a prompt injection with a handshake.

AdvancedTOOLSClaude Code

Why it works

The connector catalog is a storefront; the MCP ecosystem behind it is thousands of servers deep. Any service with an MCP server is one docs-link away: paste the link, tell the agent to wire it up and verify, and it edits its own configuration. The gate that matters isn't technical — it's trust, because a malicious server is a prompt injection with a handshake.

How to do it

  1. Find the MCP server's documentation or repository page for the service you need.
  2. Vet the author first: official vendor? well-known maintainer? recently updated?
  3. Paste the link with the instruction: 'connect this MCP server, figure out the configuration yourself.'
  4. Require proof: 'then call one of its tools and show me it actually works.'
  5. The agent edits its config, restarts what's needed, and demonstrates a live call.
  6. Record it in CLAUDE.md if project-specific: 'this project uses the X server for Y.'

Copy this

prompt
Connect this MCP server: [docs link]. Work out the configuration yourself, install it, then prove it works by calling one of its tools and showing me the result.

What to watch for

  • A server from a random forum post can exfiltrate what your agent can see — the browser-extension rule applies verbatim: source matters more than features.
  • If the agent can't make a test call succeed, remove the config rather than shipping a half-wired server that fails at 6pm on Friday.

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 →