Roadmap for Open Agentic 2.0 (v0.2.0 → v1.0.0)

[PAXECT-Interface]

Roadmap for Open Agentic 2.0

This is a rough, living roadmap for Open Agentic 2.0.
Nothing here is a hard promise, but it shows the direction the project is
currently aiming for.

v0.2.0 — Hardening the core

Focus: better tests, clearer examples, and making the micro-orchestrator easier
to adopt in small projects.

Planned items (subject to change):

• Add more tests around failure modes:
  • plugin timeouts,
  • malformed / non-JSON responses,
  • missing or low-quality evidence.
• Improve examples for plan.json, policy.yaml, and plugins.yaml.
• Add basic docs: "How to embed Open Agentic 2.0 into your own project".

v0.3.0 — Visibility & integrations

Focus: making it easier to inspect audit chains and plug Open Agentic 2.0 into
other tools.

Planned ideas:

• A simple CLI or web viewer for audit chains (plain JSONL in, human-friendly
  view out).
• More real-world policy/plan examples (different budgets, tools, and failure
  modes).
• Example integrations with external agents or systems via the legacy and
  meta plugins.

v1.0.0 — Stability & long-term use

Focus: treating Open Agentic 2.0 as a stable building block for long-lived
systems.

Possible goals:

• A stable, well-documented CLI and config surface.
• Clear, documented key management patterns for HMAC-based audit chains.
• Guidelines for running this in production pipelines (logging, retention,
  monitoring).
• A simple long-term support story (e.g. maintained 1.x branch).

---

🤝 How to influence the roadmap

Nothing here is “locked in”. If you have ideas, needs, or use cases that don’t
fit this roadmap, please:

• open a new Discussion, or
• comment here with:
  • what you’re building,
  • what you need from an audit-first agent core,
  • and how you’d like to contribute (code, tests, design, docs).

The goal is for Open Agentic 2.0 to stay small, auditable, and community-driven
— not to become a giant monolith.