Skip to content

ROADMAP.md

Latest commit

 

History

History
93 lines (70 loc) · 4.1 KB

ROADMAP.md

File metadata and controls

93 lines (70 loc) · 4.1 KB
Raw

Gobby Roadmap

Gobby is a local-first control plane for AI coding tools: persistent sessions, task graphs, workflows, hooks, MCP proxying, agents, memory, and deterministic automation around the tools developers already use.

Last refreshed: May 24, 2026.

The data migration is complete. PostgreSQL is the runtime hub, FalkorDB is the graph backend, and .gobby/tasks.jsonl remains the git-native task projection. The roadmap now tracks the post-migration release line.

0.5.0 - New Baseline

0.5.0 is the new baseline release after the PostgreSQL and FalkorDB cutover.

  • Ship the current Python daemon as the supported local-first runtime.
  • Treat PostgreSQL as the only runtime hub and FalkorDB as the supported graph backend across daemon, web UI, admin payloads, setup, and docs.
  • Keep .gobby/tasks.jsonl as the checked-in task sync projection for reviewable, task-linked commits.
  • Tighten operator docs, install/status output, and release notes around the new storage baseline.

0.5.0+ - UI Hardening And Rust Port Work

After the baseline release, the main work is web UI hardening and preparing the Rust port.

  • Harden chat, sessions, tasks, workflows, cron, projects, compact layouts, and shared design tokens until the web UI is solid enough for daily Gobby Pro use.
  • Close attached-session parity gaps: context usage, mode/model sync, attachments, persona switching, STT/TTS, and first-class web chat behavior.
  • Continue plan registry APIs and UI editors so stage and build-profile shape can evolve without hand-editing storage or YAML internals.
  • Finish logging cleanup before enforcing logging-format rules: config reset, runtime-vs-app log separation, normalized handlers, automation logs for cron and dispatch, and quieter routine logging.
  • Prepare the Rust port by freezing route contracts, adding compare/delegation plumbing, and extracting shared primitives in gobby-cli.

0.6.0 - Rust Port Release

0.6.0 is the Rust port release. It is an incremental strangler port, not a rewrite.

  • Python remains the public daemon and behavioral reference until each boundary passes parity, observability, and rollback gates.
  • Rust sidecars run on internal ports, with Python delegating selected route families behind explicit flags.
  • Compare mode calls both implementations and returns the Python response until parity is proven.
  • The bridgehead stays in gobby-cli: gcode, gsqz, gloc, ghook, and gobby-core shared primitives.

0.6.0+ - Gobby Pro Sync

Gobby Pro starts with remote sync from multiple Gobby-controlled machines.

  • Multi-daemon discovery and handshake.
  • Opt-in encrypted sync for tasks, memories, and session metadata.
  • Operator controls for machines you own, with local-first behavior preserved.
  • Sync conflict handling, audit trails, and release packaging for the commercial layer.

0.7.0+ - Gobby Pro Beta

The Pro beta introduces fleet management and a shared dashboard for all Gobby machines.

  • Fleet inventory, health, and remote command.
  • Shared task boards, team workflows, and review state across machines.
  • Dashboard views for sessions, builds, agents, validation, and sync health.
  • Enterprise controls for audit, policy, and team operations.

Later - 1.0.0 Prep

Once the Pro beta stabilizes, the focus moves to 1.0.0 readiness.

  • Stabilize public APIs, configuration, workflow definitions, and hook behavior.
  • Polish install, upgrade, recovery, and operator documentation.
  • Add SWE-bench evaluation from docs/plans/SWE-BENCH.md: eval run/result storage, gobby eval, Docker-backed harness, trajectory capture, leaderboard artifacts, score tracking, and Gobby-enabled vs baseline A/B tests.
  • Use the benchmark results to drive final release gates for 1.0.0.

Baseline Already Shipped

0.4.x shipped the task-to-PR loop as the supported path: persistent sessions, cross-CLI handoffs, task lifecycle and validation, MCP progressive discovery, workflows, pipelines, rule enforcement, agent spawning, memory search, skills, integrations, and the web UI surfaces needed to operate Gobby locally.