AI sandbox runtime — caging autonomous agents at ring-0
securityrustsandboxai
- r syscall: deny ptrace from sandboxed processesrich · 16 Dec
b02953bView →
- r sandbox: ring-0 caging for autonomous agentsrich · 14 Dec
4eb6d23View →
- r tests: regression — escape via clone+namespacerich · 13 Dec
a639ee7View →
- r docs: threat model — when ring-0 is sufficientrich · 10 Dec
319ebaeView →
- r kernel: trim BPF surface for sandbox helpersrich · 8 Dec
5215661View →
- r fix: race in seccomp filter installationrich · 27 Nov
00df94cView →
- r audit: enable lockdown=integrity at bootrich · 26 Nov
29b4b21View →
- r refactor: split policy engine into core + clirich · 23 Nov
07decd1View →
- r feat: per-agent capability allowlistrich · 20 Nov
a7da70fView →
- r fix: file descriptor leak on policy reloadrich · 15 Nov
a462121View →
- r docs: explain why we picked landlockrich · 14 Nov
cd52d89View →
- r chore: vendor microhttpd at 0.9.78rich · 10 Nov
f781865View →
- r tests: fuzz seccomp filter with afl++rich · 8 Nov
5f87258View →
- r syscall: emulate getrandom for sandboxed jailsrich · 6 Nov
3c32538View →
- r feat: trace mode with verbose syscall loggingrich · 4 Nov
e5715adView →
- r perf: avoid copy on syscall arg inspectrich · 2 Nov
b5b68eaView →
- r fix: kill agent on ENOMEM instead of OOM oom-killrich · 30 Oct
0c2aaa4View →
- r docs: prod deployment hardening checklistrich · 27 Oct
afe01f5View →
- r refactor: rename SandboxPolicy → CageSpecrich · 26 Oct
0102e30View →
- r feat: ring-0 LSM hooks for agent introspectionrich · 22 Oct
afecae7View →
- r docs: white paper draft v1rich · 18 Oct
466d2dbView →
- r kernel: harden mmap to deny PROT_EXEC by defaultrich · 14 Oct
b64b267View →
- r feat: post-exec audit — what did the agent touchrich · 8 Oct
5099709View →
- r fix: stack-clash mitigation via guard pagesrich · 6 Oct
7ea0d28View →
- r perf: micro-optimise syscall hot pathrich · 3 Oct
f03dfbeView →
- r refactor: drop dead policy dialect parserrich · 30 Sept
21e6254View →