method · first seen 5 Oct 2026
sandboxing
Also: sandbox, malware sandbox analysis, application virtualization
Running untrusted code or agents in an isolated environment to limit their access to the host system.
Sandboxing means running untrusted code in an isolated environment so it cannot freely reach the host system. The idea is decades old. Its current uses span malware analysis, application virtualization and, more recently, containing AI coding agents. The sources show this spread. One paper turns CAPEv2 malware-sandbox output into ranked threat intelligence. Another proposes object-bound confinement for compromised coding agents. A Hacker News post covers running a Docker agent in a sandbox, and an Elixir repo runs untrusted Erlang with no ambient authority. Much of the GitHub signal is noise: four near-identical 'ai--manus' forks posted on the same day, plus an SEO-style download repo.
Agent-specific confinement work, such as binding permission checks to objects rather than names, could become standard practice as coding agents act with user authority. Sandboxing as a general concept, however, is long established, and the measured rise here is inflated by duplicate repositories.
Analyst note (established), 17 Nov 2026.
Mentions per week
PapersCodeDiscussion
Rank 104 on 14 Nov 2026 · 6 mentions in 28 days · 6 independent groups · papers → discussion → code · first seen 6 weeks ago · acceleration +1.79
Earliest sources
- From Sandbox to Enforcement: Confidence-Qualified Threat Intelligence for Critical InfrastructurePapers · 5 Oct 2026 · Nikolaos Kekatos, Mihaela Curcă, Georgios Koutidis, Mihai Nena, Tom Nianios, Ro…
- Zed Editor, Docker Agent, ACP, but in a sandbox: running the agent with sbxDiscussion · 7 Oct 2026 · speckx · 25 points
- Names Are Not Objects: Object-Bound Confinement for Compromised AI Coding AgentsPapers · 10 Oct 2026 · Daniel Commey
- hyperbob-labs/hyperlangCode · 26 Oct 2026 · hyperbob-labs · 34 stars
- nageshbhaskar2941-boop/shadow-defenderCode · 10 Nov 2026 · nageshbhaskar2941-boop · 102 stars
- TorusBarbarianMap9/ai--manusCode · 13 Nov 2026 · TorusBarbarianMap9 · 187 stars
Latest sources
- ttobulevicius-cpu/ai--manusCode · 13 Nov 2026 · ttobulevicius-cpu · 187 stars
- PolygonGlazier/ai--manusCode · 13 Nov 2026 · PolygonGlazier · 187 stars
- ArcticPrawnEmporium/ai--manusCode · 13 Nov 2026 · ArcticPrawnEmporium · 187 stars
- TorusBarbarianMap9/ai--manusCode · 13 Nov 2026 · TorusBarbarianMap9 · 187 stars
- nageshbhaskar2941-boop/shadow-defenderCode · 10 Nov 2026 · nageshbhaskar2941-boop · 102 stars
- hyperbob-labs/hyperlangCode · 26 Oct 2026 · hyperbob-labs · 34 stars
- Names Are Not Objects: Object-Bound Confinement for Compromised AI Coding AgentsPapers · 10 Oct 2026 · Daniel Commey
- Zed Editor, Docker Agent, ACP, but in a sandbox: running the agent with sbxDiscussion · 7 Oct 2026 · speckx · 25 points
- From Sandbox to Enforcement: Confidence-Qualified Threat Intelligence for Critical InfrastructurePapers · 5 Oct 2026 · Nikolaos Kekatos, Mihaela Curcă, Georgios Koutidis, Mihai Nena, Tom Nianios, Ro…