✍️Writing & Content23🎨Image Generation32🎬Video & Animation64🎵Audio & Music49💬Chatbots & Assistants36💻Coding & Development155📈Marketing & SEO52Productivity139🎯Design & UI/UX52📊Data & Analytics36📚Education & Research24💼Business & Finance50🏥Healthcare & Wellness18🔍Search & Knowledge14🤖AI Agent Infrastructure25🛡️AI Security & Testing2🧊3D & Spatial14🔎SEO Tools5🏡Real Estate4🗃️Data Extraction3🧠ADHD & Focus Tools9
Listed in AI Agent Infrastructure with 28 other toolsPart of 940+ curated AI tools on AISO
Tilde logo

Tilde

Transactional agent sandbox — versioned filesystem, atomic rollback, default-deny egress, full audit trail

freemiumCredit-based. Free: 1,000 credits/month, 1 seat, 3 concurrent sandboxes, 5 GB storage, 25 GB egress, hard caps so no bill is possible. Pro $49/month: 8,500 credits, 3 seats (+$10/seat), 10 concurrent sandboxes, 200 GB egress, one month credit rollover, 48-hour email support, overage at $0.012/credit.View full pricing →

About Tilde

Tilde is a sandbox runtime that makes it safe to point autonomous agents at real production data by turning every agent run into a transaction you can roll back. Code from GitHub, objects from S3, and documents from Google Drive are mounted as a single versioned POSIX filesystem at ~/sandbox — a real filesystem, so any tool in any language works without an SDK — and every file is versioned from the first commit, meaning any run can be reverted with one command instead of restored from a backup. Each run executes in a fresh isolated container: on a clean exit changes commit atomically, and on failure nothing changes at all, so there is no half-written state to clean up. Network egress is default-deny and policy-checked per request, with cloud metadata endpoints and private ranges blocked outright, which closes off the data-exfiltration and prompt-injected-callout failure modes directly rather than hoping the model behaves; every allowed and denied request is logged against the agent that made it. On top of that sits a time-travel audit trail showing exactly what changed, who did it, and when, down to the individual file, plus per-action policies and human approval gates for operations you want a person to sign off on. It integrates with Claude, LangGraph, Hugging Face, S3, and Google Drive.

Key Features

GitHub, S3, and Google Drive mounted as one versioned POSIX filesystem
Every agent run is an atomic transaction — clean exit commits, failure reverts
One-command rollback of any run
Default-deny network egress with per-request policy checks and logging
Time-travel audit trail down to the file
Per-action policies and human approval gates

Tags

agent sandboxsecurityversioned filesystemauditrollbackegress control
🏷️

Is this your tool?

Claim your listing to get a Featured badge, edit your description, and stand out from competitors. All plans include a permanent dofollow backlink to your site.

Claim Now →

ChatGPT already recommends Tilde. Does it recommend yours?

If you're building in AI Agent Infrastructure, run a free AI-visibility scan on your own product — we ask ChatGPT across 5 prompt angles and score how often you get named. ~30 seconds, no signup, no card.

Stay updated on AI Agent Infrastructure tools — join our weekly newsletter

One concise email with fresh launches, trending picks, and featured standouts.

Alternatives to Tilde

View all Tilde alternatives →

Agent connectivity: not yet verified