Skip to content

Agno AgentOS API

AI Agents Agno 3.1.11 Hardened AMI

Agno 3.1.11 on Ubuntu 24.04 serves a CSV analyst through a local authenticated API and keeps conversation sessions in SQLite. You bring an OpenAI-compatible model that can call tools and return a non-streaming chat completion. The image does not include model weights, a commercial Agno control-plane UI, hosted Agno, or SSO.

AMI release
0.1.2-r20261105 (Agno 3.1.11)
Recommended
t3.medium
Storage
24 GiB+ encrypted gp3
Access
Session Manager, no inbound ports
Released
Nov 5, 2026
Delivery
Amazon Machine Image in your own AWS account

Local authenticated AgentOS API with a native Agno CSV calculation tool and independently checkable JSON results.

SQLite conversation sessions and task files retained across idle service restarts, with documented backup and recovery.

Bring an OpenAI-compatible provider with tool calling; no bundled model weights or commercial control-plane UI.

Quick start

Full guide →
  1. Subscribe
  2. Launch the instance
  3. Connect
  4. Configure a model
  5. Run the CSV task
  6. Read it again after a restart

What is on the image

This guide matches Marketplace version 0.1.2-r20261105 (Agno 3.1.11). The public customer guide describes package 0.1.2 (Agno 3.1.11). Check the listing for the version currently on offer.

  • Agno 3.1.11 under /opt/agno/. The service can write only /var/lib/agno/.
  • Loopback API at 127.0.0.1:7777. Every route, including health, requires a Bearer token. There is no browser UI. Browser Origin requests are rejected.
  • One CSV analyst and a calculation tool. No shell, email, Slack, or CRM tools are configured. Agno telemetry and OpenTelemetry collection are disabled.
  • SQLite sessions in /var/lib/agno/sessions.db, and task files under /var/lib/agno/workspace/.
  • Helpers sushi-agno and /opt/sushi/docs/demo.sh.

t3.medium (4 GiB of memory) or a larger listed type, with an encrypted gp3 root volume of at least 24 GiB and IMDSv2. No GPU is required. The listed types are t3.medium, t3.large, t3.xlarge, and t3.2xlarge.

Scope and limits

The calculation tool accepts at most 1,000 rows and 64 KiB, and writes result.json once. Model requests have a 60-second timeout, no SDK retries, and a 512-token response limit. The local API token is an administrator credential for this single-user runtime, including upstream session APIs. It is not a multi-tenant service.

Unresolved operating-system findings

The 0.1.2-r20261105 release note says known OS and kernel High and Critical vulnerability findings remain unresolved, and that this delivery does not claim a security fix or a security qualification.

Billing

The listing states USD 0.03 per running instance-hour. Confirm the offer, including its free trial, before you subscribe. EC2, EBS, data transfer, and model-provider usage are billed separately. The included demo normally makes two model requests.

Where to go next

Upstream project

Agno 3.1.11 is packaged under Apache-2.0. Software Sushi does not claim Agno or AWS endorsement. Notices are in /usr/share/doc/agno/. Original source and wheel inputs are in /opt/agno/sources/ and /opt/agno/wheels/. See the Agno documentation. Do not upgrade the installed environment in place with an unpinned pip command.

Last updated November 11, 2026 · Suggest an improvement