Capabilities
Honest status of what owlcoda can do today. We distinguish between supported, partial, and unsupported — no ambiguity.
Status definitions
- Supported — Works reliably in current version. Tested and usable.
- Partial — Works in some cases or with limitations. May require manual steps or specific conditions.
- Unsupported — Not available yet. Planned or under consideration, but not functional today.
Context & Continuity
| Capability | Status | Notes |
|---|---|---|
| Structured project context creation | Supported | Create structured context from scratch or from existing materials. |
| Context persistence across sessions | Supported | Project context survives session boundaries via file-based storage. |
| Context recovery on return | Supported | Resume work with full project awareness. No re-explanation needed. |
| Automatic context summarization | Partial | Works for most project sizes. Very large contexts may need manual curation. |
| Real-time context sync (multi-user) | Unsupported | Currently single-user. Multi-user sync is not yet implemented. |
Handoff
| Capability | Status | Notes |
|---|---|---|
| Session-to-session handoff (same user) | Supported | Close a session and resume later with full context. |
| Model-to-model handoff | Supported | Switch between AI models while preserving project context. |
| Person-to-person handoff | Partial | Works via shared file access. No built-in user management or permissions yet. |
| Automated handoff generation | Partial | Handoff documents are generated but may need human review for accuracy. |
| Handoff via API | Unsupported | No API endpoint yet. Handoff is file-based. |
Review & Execution
| Capability | Status | Notes |
|---|---|---|
| Project state review | Supported | Inspect completed, in-progress, and pending work items. |
| Structured execution tracking | Supported | Actions are tracked within project context, not just chat messages. |
| Phase reset | Supported | Close out a phase and start fresh while preserving history. |
| Dependency tracking between work items | Partial | Basic dependency tracking works. Complex dependency graphs need manual management. |
| Automated execution (agent mode) | Unsupported | owlcoda does not autonomously execute tasks. A human or model must drive execution. |
Multi-Model Support
| Capability | Status | Notes |
|---|---|---|
| Works with multiple AI providers | Supported | Context is model-agnostic. Works with any model that can read structured text. |
| Mid-project model switching | Supported | Switch models between sessions without context loss. |
| Parallel multi-model execution | Unsupported | Using multiple models simultaneously on the same project is not yet supported. |
| Model-specific context formatting | Partial | Basic formatting works across models. Provider-specific optimizations are limited. |
Infrastructure & Integration
| Capability | Status | Notes |
|---|---|---|
| Local file-based storage | Supported | All data stored locally. No cloud dependency for core functionality. |
| Git integration | Partial | Works alongside git repos. No deep git integration (auto-commit, branch management) yet. |
| Cloud storage / sync | Unsupported | No cloud storage option. All context is local. |
| Web UI | Unsupported | No web interface. owlcoda currently operates through text-based context files. |
| IDE / editor integration | Unsupported | No editor plugins yet. Planned for future versions. |
Summary
owlcoda v0.1 is a working system for structured work continuity. Its core strengths are context persistence, session-to-session handoff, model-to-model portability, and structured review/execution tracking.
It is not a web app, not a cloud service, and not an autonomous agent. It's a local, file-based continuity layer that works with whatever models and tools you already use.
Capabilities will expand. This page will be updated as features move from unsupported to partial to supported.