Current roadmap and known limitations

[huangruiteng]

This thread summarizes the current public direction. It is a maintainer-curated roadmap, not a delivery promise or a substitute for shipped release notes.

Current baseline

The v0.4.x line is an early but usable local control plane for long-running agent work. It ships durable goals and state, typed todos and decision scopes, peer claims and leases, evidence and writeback, quota-aware scheduling, cross-turn continuation, public validation canaries, and read-first operator surfaces.

Near-term roadmap

• simpler installation, release verification, and host packaging;
• broader typed runtime and host adapters;
• stronger repeated-loop terminal acceptance and recovery;
• independent adoption and outcome evidence;
• clearer management, review, and community surfaces.

Known limitations

• LoopX is not an agent runtime, a full agent platform, or an autonomous production controller.
• Advanced host integrations, Auto Research paths, and domain-specific capabilities may be optional, default-off, or experimental.
• LoopX does not grant credentials, approve destructive or production actions, or convert an unverified run into evidence of success.
• Local active state and private evidence are intentionally kept outside the public repository.
• Public showcases are source-labeled. Redacted owner-run material is not a production result, company endorsement, or independently reproducible evaluation.

Follow progress

• Releases
• Open issues
• Public update notes
• Product roadmaps

Reply with the limitation that most affects your workflow, ideally with a public or synthetic example. Maintainers will update this top post when the baseline materially changes.