Questions about Grepticon
What is Grepticon?
Grepticon provides a hosted filesystem service for AI agents to perform retrieval operations over isolated, permissioned workspaces. It offers four read tools—ls, find, cat, and grep—that operate over a real filesystem without requiring a sandbox to provision or embeddings to maintain. The platform supports per-tenant token scoping and maintains an append-only audit log of all read activity, accessible via a TypeScript SDK, the Vercel AI SDK, a REST API, and an MCP adapter.
Who is Grepticon for?
The service is designed for teams shipping agents to production who need retrieval over isolated, permissioned workspaces without the overhead of a vector database. It is accessible via a TypeScript SDK, the Vercel AI SDK, a REST API, and an MCP adapter, making it suitable for developers integrating AI into their documentation pages in minutes.
How can I get started or book Grepticon?
Users can start building immediately with a free plan that includes three workspaces, 500 MiB of storage, and 10,000 files per workspace. The Pro plan costs $10 per month and offers 100 workspaces and 5 GiB of storage, while Enterprise plans provide custom limits and white-glove onboarding for organizations with specific compliance needs.
What counts against the read rate limits?
The platform's read rate limits only include the four read tools: ls, find, cat, and grep. Creating workspaces, uploading files, and minting tokens never count against these limits, ensuring that core retrieval operations remain fast and predictable for agents.
What happens when I hit a limit?
When a workspace or account hits a limit, the system returns a 429 status with a Retry-After header rather than silently dropping writes. The SDK automatically retries these requests, and quotas fail the write with a clear message to prevent data loss.
What makes Grepticon different from competitors?
Grepticon differentiates itself by offering a filesystem that beats a vector database for agent retrieval, scoring 16.7 points higher than naive RAG across an 18,340-file corpus. The benchmark showed 97.9% retrieval accuracy, significantly outperforming both naive RAG and agentic-RAG methods.
How does access control and auditing work?
The service maintains an append-only audit log that records which token, which verb, which argument, and what came back for every read operation. Users can filter or export this log to ensure full visibility into all retrieval activities within their workspaces.
How does token scoping work?
The platform supports per-session tokens scoped to one tenant, ensuring that what a token cannot reach is absent from every listing, glob, and grep hit. This token scoping mechanism prevents unauthorized access to files outside the defined permissions.
What tools and integrations are available?
The service is accessible via a TypeScript SDK, the Vercel AI SDK, a REST API, and an MCP adapter, allowing developers to integrate it into their existing workflows quickly. Additionally, a Soongrepticon CLI is available for command-line interactions.
What are the pricing tiers and what happens when you reach them?
The pricing structure includes a Free tier with no card required, a Pro tier at $10 per month, and an Enterprise tier with custom billing. All plans run the full product, hosted and audited, with higher tiers only lifting the limits.
What are the file size limits?
The platform supports up to 50 MiB max file size across all plans, ensuring that even large documents can be processed efficiently by the read tools. This limit applies consistently from the Free plan up to the Enterprise tier.
How can I learn more or get support?
The platform provides a quickstart guide and documentation to help users integrate AI into their docs page in minutes. It also offers a support agent for assistance with onboarding and technical questions.
How does the platform handle different backends?
The service is designed to work with various backends through a single interface, allowing flexibility in how agents interact with the filesystem. This approach ensures compatibility across different environments without requiring schema or framework glue.
What does the audit log record?
The platform's audit log records every read operation, including the token, verb, argument, and response, providing a complete trail of all retrieval activities. This feature is essential for compliance and security requirements in enterprise environments.
What is the official website of Grepticon?
The official website of Grepticon is grepticon.com.
Post your question publicly — other visitors, and whoever manages this page, can answer. It will be visible on this page.
Want to hear about this company again? Follow it and we’ll email you when a new review appears.
Based on reviews, listing data and the company's website. Verify anything important with the company directly.
Questions & Answers
How do the ls, find, cat, and grep tools access the underlying files without needing a sandbox?
Trustburn Team suggests a starter question on new profiles — the business hasn’t answered yet.
Sign in to post
Your text is saved — you'll come right back to post.
Don't have an account? Create one
News & Press
No news yet. This is where the business shares its own updates and press.
Photos
No photos yet. Photos shared by reviewers will show up here.
About Grepticon
Grepticon provides a hosted filesystem service designed for AI agents to perform retrieval operations over isolated, permissioned workspaces. Its platform offers four read tools—ls, find, cat, and grep—that operate over a real filesystem without requiring a sandbox to provision or embeddings to maintain. The service supports per-tenant token scoping and maintains an append-only audit log of all read activity. It is accessible via a TypeScript SDK, the Vercel AI SDK, a REST API, and an MCP adapter.
- Website
- grepticon.com
- [email protected]
