Pantheon Guard · alpha
Already run Claude Code? It has your shell.
One command puts your existing coding-agent sessions on the same tamper-evident record — and gives them an "ask me first." Dangerous commands — rm -rf, force pushes, sudo, deploys — pause the session and buzz your phone. One tap decides.
- ✓Everything recorded — never breaks your session
- ✓Only the dangerous is gated — no approval fatigue
- ✓Your org's policy rules apply — auto-allow the safe, deny the forbidden, gate the rest
- ✓Fail closed — gate unreachable means blocked, never silently allowed
No migration, no new agent to adopt. Works with the agents you already trust with your code.
$ npx @pantheon-os/guard init
✓ Pantheon Guard installed.
⏺ Bash(git push origin main --force)
⧗ Pantheon Guard: "force push" needs approval — check your phone…
📱 claude-code session
[force push] git push --force
✓ Approved — proceeding.
⛓ recorded on your chain · seq 1,204
The gating and ingestion API is live and documented today. The one-command installer publishes to npm shortly — design partners get it directly.
Build the operating layer for consequential AI work. Start with one provable release.
Pantheon's mission spans the organization. The current founder-led pilot proves that mission on one GitHub-backed release path, one success metric and one decision-to-outcome receipt.
Starts outside production. Production follows only after T‑09/T‑10 and jointly agreed rollback conditions pass.