Skip to content

Future Implementation Roadmap

This roadmap is forward-looking and organized into three delivery phases.

Phase 1 - Short-Term (Next 1-3 Months)

Milestones and Deliverables

  1. Runtime reliability hardening
  2. Deliverables:
  3. improved startup diagnostics for provider and MCP failures

  4. clearer degraded-mode notifications in Telegram and dashboard

  5. Documentation completeness and consistency

  6. Deliverables:
  7. finalized doc index and per-folder docs
  8. configuration examples for common provider combinations

  9. operator runbooks for incident triage

  10. Scheduler usability improvements

  11. Deliverables:
  12. richer natural-language schedule parsing coverage
  13. better validation/error messages for /schedule add workflow payloads

Target Enhancements

  • Expand tests around autonomous sandbox loop guardrails.
  • Add explicit docs for fallback pathways and recovery playbooks.

Phase 2 - Medium-Term (3-6 Months)

Major Features

  1. Multi-channel interface expansion

  2. Add additional chat channels beyond Telegram with shared memory semantics.

  3. Enhanced memory governance

  4. Add retention, archival, and user-directed memory controls.

  5. Advanced dashboard operations

  6. Role-aware admin views and richer historical analytics.

Refactoring Tasks

  • Split large Telegram handler module into domain-oriented submodules.
  • Extract scheduler workflow evaluator into dedicated component with tighter typing.
  • Consolidate duplicated runtime paths (her-core/main.py and legacy her-core/telegram_bot.py) behind one supported entry strategy.

Stability and Performance

  • Improve response latency under provider failover.
  • Add load-focused runtime checks and memory query profiling.

Phase 3 - Long-Term (6-12 Months)

Visionary Features

  1. Adaptive policy engine

  2. Policy-driven behavior tuning with explicit safety and style contracts.

  3. Plugin ecosystem for tools

  4. Structured extension model for MCP profiles and custom tools.

  5. Explainability layer

  6. Human-readable rationale summaries for selected decisions and actions.

Platform Expansion

  • Broader deployment targets (managed container platforms, autoscaling profiles).
  • More provider presets and memory backend options.

Scaling and Community

  • Contributor templates for modules/tests/docs.
  • Public issue labels and RFC workflow for architecture-level changes.
  • Versioned documentation and migration guides for major releases.