Skip to content
Back to search

ranwhat

MarketplaceWorks with Claude CodeGitHub

Audit log and secret scanner for AI coding agents: see what Claude Code, Codex, Gemini CLI and 27 more ran on your machine, and find API keys leaked into their transcripts. Local, MIT.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/matijamiki--ranwhat/badge.svg)](/listings/matijamiki--ranwhat)

HTML

<a href="/listings/matijamiki--ranwhat"><img src="/api/listings/matijamiki--ranwhat/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Try it once without installing anything:

uvx ranwhat check

To keep it, install with `pip install ranwhat`, `uv tool install ranwhat` or `brew install MatijaMiki/ranwhat/ranwhat`. Needs Python 3.9 or later.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 15 lines
  1. {
  2. "name": "ranwhat",
  3. "owner": { "name": "Matija Mikulinec" },
  4. "metadata": {
  5. "description": "ranwhat: see what your AI coding agents ran, and what secrets they left behind."
  6. },
  7. "plugins": [
  8. {
  9. "name": "ranwhat",
  10. "source": "./plugins/ranwhat",
  11. "description": "Flight recorder and secret finder for AI coding agents. Reads local transcripts, sends nothing.",
  12. "homepage": "https://ranwhat.com"
  13. }
  14. ]
  15. }
README.md · README · 93 lines
  1. # ranwhat
  2. ranwhat checks whether any of your secrets have leaked into your coding
  3. agents' history, and where. It also checks whether your agent ran anything
  4. sketchy: read your SSH key, force-pushed, piped a script from the internet
  5. into a shell, published a package.
  6. It reads the history that Claude Code and twenty-nine other coding agents
  7. already keep on your disk, including Codex, Gemini CLI, Cursor and GitHub
  8. Copilot
  9. ([full list](https://github.com/MatijaMiki/ranwhat/blob/main/docs/reference.md#ranwhat-watch-audit-what-claude-code-and-your-other-agents-ran)).
  10. It's open source (MIT), runs locally and has no dependencies. No account
  11. needed, no telemetry.
  12. Use it as an audit log for your agents after the fact: see what commands
  13. Claude Code ran on your machine, and which API keys and passwords ended up
  14. in its transcripts.
  15. ## Install
  16. Try it once without installing anything:
  17. ```bash
  18. uvx ranwhat check
  19. ```
  20. To keep it, install with `pip install ranwhat`, `uv tool install ranwhat` or
  21. `brew install MatijaMiki/ranwhat/ranwhat`. Needs Python 3.9 or later.
  22. ## Quick start
  23. ```bash
  24. ranwhat check
  25. ```
  26. It changes none of your files. It lists the risky things your agents ran in
  27. the last 30 days, and the secrets it finds in those transcripts and where
  28. each one was found (`--days 365` looks further back):
  29. ```
  30. ranwhat watch · local agent flight recorder
  31. ...
  32. * Credential material accessed 2026-09-30 18:13:00 Bash 9c41e07b2d5a8f13
  33. cat ~/.ssh/id_rsa
  34. -> The agent read a file whose only purpose is to hold secrets. Whatever
  35. it read is now in a model context you do not control.
  36. * Destructive git operation 2026-09-15 17:30:00 Bash e181acd29cf199ab
  37. git push --force origin main
  38. -> History rewriting or branch deletion. This is the class of action
  39. that destroys the record of what else happened.
  40. ranwhat clean · secrets sitting in local transcripts
  41. ...
  42. * Stripe live secret key sk_…dc 32 chars seen 8x
  43. read from api/.env
  44. in /Users/you/Desktop/app
  45. ```
  46. Then `ranwhat clean` walks you through the secrets it found. You can mask
  47. the ones in plain transcript files, see what to rotate, and open each
  48. provider's page to do it. Agents that keep a database, like Cursor, are only
  49. read, and it tells you how to remove the secret there. Masking stops a key
  50. leaking a second time but doesn't un-leak it, so rotate it.
  51. If you want Claude Code to ask you before a high-risk command runs, there's
  52. an opt-in hook: `ranwhat hook install`.
  53. To keep a record of what `watch` flagged that shows if it was changed later,
  54. `ranwhat ledger record` appends it to a hash chain on your machine, and
  55. `ranwhat ledger verify` names the first line that no longer fits.
  56. ## Free and paid
  57. Everything that runs on your machine is free and stays free. Plus is €12 a
  58. month per organisation, for the things that need a server: findings from
  59. all your machines and CI runners in one place (`ranwhat push`), CI tokens
  60. and teammates on your account, and a catalogue of what each token scope
  61. allows, kept up to date. See [pricing](https://ranwhat.com/pricing).
  62. ## More
  63. - [ranwhat.com](https://ranwhat.com): docs and guides
  64. - [Commands](https://ranwhat.com/commands): every command and flag
  65. - [GitHub Action](https://github.com/MatijaMiki/ranwhat/blob/main/docs/reference.md#in-ci-the-github-action):
  66. run `ranwhat check` in CI after the steps where an agent ran; it can fail
  67. the job on findings and uploads SARIF
  68. - [Full reference](https://github.com/MatijaMiki/ranwhat/blob/main/docs/reference.md):
  69. where each agent keeps its history, the rules, the ignore file, the Claude
  70. Code plugin and exactly what goes online
  71. - [Issues](https://github.com/MatijaMiki/ranwhat/issues)
  72. MIT licence. Built by cenner.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.