# Roadmap

Canonical HTML: https://www.praxis-agents.ai/roadmap

Last updated: 10 July 2026

See what is available now, what is being built next, and which safety and governance foundations come before higher-risk capabilities.

## Available now

- agent runtime, conversations, streaming, delegation, schedules, and cancellation
- workspace identity, roles, invitations, approvals, audit, and side-effect envelopes
- typed tool registry, audited dispatch, skills, and skill documents
- file storage, revisions, processing, agent file tools, file UI, and multimodal input
- transport retries, token caps, tracing hooks, quality gates, and runtime hardening

## Planned next and later

- integration credentials, OAuth, discovery, active context, and first providers
- agent behaviour evaluation, bounded tool results, and prompt-injection threat modelling
- knowledge ingestion, hybrid retrieval, and governed knowledge tools
- persistent agent memory and visible memory surfaces
- versioned artifacts, sandboxed execution, and public launch readiness

Follow [detailed milestones on GitHub](https://github.com/Greg-Asquith/praxis-agents-os/blob/main/docs/plans/000_MASTER_ROADMAP.md).
