CAIN-42 CAIN StudioAttack lab
Every button is a real agent action sent to the live gate. Edit the command and try to get it allowed.
12 attacks 2 controls live: nothing pre-recorded
- Pick a preset. Red ones are attacks, green ones are honest requests that must still work.
- Tamper. Choose any field and drop it, null it, change its type, upper-case it, flip one character, or add a field the product never asked for.
- Send and read the real reply, or run every preset and watch the scoreboard fill.
What each preset tests
| Preset | Route | Why it matters | |
|---|---|---|---|
| attack | Agent action: Install script piped straight into bash | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Agent uploads .env to a paste site | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Clean-up that deletes the home directory | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Force-push over main | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Base64-hidden command run by bash | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Destructive command hidden inside sh -c | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Agent 'fixes' networking by flushing the firewall | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Agent loosens its own permission settings | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: API key written into source code | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Agent action: Agent edits the CI workflow | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| control | Agent action: Run the test suite (safe) | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| control | Agent action: Delete the build folder inside the project (safe) | POST /codeguard-keystone/check | The same gate a coding agent's hook calls before a shell command or file write runs. |
| attack | Your command: base64-decoded payload piped to sh | POST /codeguard-keystone/check | Encoding does not hide what reaches a shell. |
| attack | Your command: $(curl) inside eval | POST /codeguard-keystone/check | Fixed in CODEGUARD 1.2.1 (Top 25 #25 blind set). |
Tamper with any field and send it again: a verifier must answer no, never crash or say yes. Public demo routes are rate limited per address; a 429 shows as rate limited. Our own lab; no independent test.
FIRME AI CORPORATION · CAIN-42 is a Byzantine trust fabric. Numbers marked "our own check" were run by us, not by an independent party. Nothing on these pages is a certification.