DEDUPED Hot AI repos, turned into use cases Get the Monday 5

How to let Codex work on a client's server with no key left behind with Agentreach

Run Codex against a remote project directory so its tools act on that server.

Who it is forA contractor or consultant who must change code or config on a client's server and does not want an agent install or API key left on a machine they do not control.

The problemAn agent installed on a client server keeps the operator's API key in env files and shell history, and that key is exposed if the server is compromised.

The outcomeCodex edits and runs commands on the client's project directory while the key stays on your own machine.

What you need first

  • Go toolchain, or a pre-built reach binary from the releases page
  • An ssh binary on your machine that already reaches the target
  • The coding agent installed on your local machine, such as Claude Code or Codex
  • Access to the target over SSH, with keys or config already set up

Steps

  1. Install reach with go install github.com/bojieli/agentreach/cmd/reach@latest
  2. Start the session with reach client-box:/srv/app codex
  3. Ask Codex to make the change and let it run its commands on the server
  4. Check the record of commands and writes with reach log
  5. Close the session with reach down using the session name from reach status

Prompt for your agent

Use the repo bojieli/agentreach to run Codex on <target>, for example client-box:/srv/app. Ask it to list the project files and report the current config value you name, then confirm the work ran remotely with `reach log`. Expected output: the file listing and config value from the server, plus log entries for the exec and write actions.

Caveat. Codex is described as the clean fit, but the README says it has been verified end-to-end only at version 0.148.0, so other versions may behave differently.

Share: X LinkedIn Hacker News Email Copy link

Disclaimer

Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page