[Roadmap] Portfolio research and execution roadmap

Goal

Build a reproducible path from dynamic universe selection through portfolio research, pair/spread modeling, execution, and auditable analytics.

This issue is the public index for the first roadmap theme. It describes direction, not a release-date promise. Implementation work belongs in scoped child issues and pull requests.

Workstreams

• #251 — Dynamic universe and point-in-time stock screening
• #252 — Portfolio strategies and scheduled rebalancing
• #253 — Pairs and spread trading foundation
• #254 — Multi-leg order coordination and recovery
• #255 — Pair-level positions, PnL, and analytics
• #256 — External signals and auditable AI entry filters

Sequencing

1. Agree point-in-time universe and portfolio lifecycle contracts.
2. Define pair/spread identity and persistence.
3. Build durable multi-leg execution and recovery.
4. Add reconciled pair analytics and user interfaces.
5. Add external signal gates using the same point-in-time guarantees.

Contribution policy

• Discuss architecture in the relevant epic before opening a large pull request.
• Only scoped work marked Ready and help wanted should be implemented without another design round.
• Comment with the task you want to own and a short implementation plan.
• A maintainer confirms the scope and assigns the issue.
• Post a draft pull request or progress update within seven days after assignment. If no update is posted, the task may be released.
• Keep pull requests focused; do not implement an entire epic in one change.

See ROADMAP.md and CONTRIBUTING.md.