Skip to content

Project Roadmap

This roadmap organises planned features into versioned milestones. Items marked [done] are already implemented in the current codebase.

v0.2 — Current release (2026-Q1)

Theme: Solid foundation — design patterns, full test suite, documented API.

  • [x] Refactor to src/ package layout (tot_agent)
  • [x] Strategy pattern for cover sources (OpenLibrarySource, GoogleBooksSource)
  • [x] Observer pattern for agent events (ConsoleObserver, LoggingObserver)
  • [x] Template Method pattern for goal builders
  • [x] Sphinx-compatible docstrings throughout
  • [x] PEP 8 compliance
  • [x] Full unit test suite (pytest + pytest-asyncio + respx)
  • [x] Integration test suite with Playwright
  • [x] Coverage reporting (pytest-cov)
  • [x] pyproject.toml with hatchling backend
  • [x] Full-featured CLI: create, vote, simulate, seed, goal, users, info, covers
  • [x] MkDocs Material docs with Mermaid diagrams
  • [x] IEEE 830 SRS
  • [x] Software design document
  • [x] Virtual environment setup

v0.3 — Robustness & configurability (2026-Q2)

Theme: Make it easier to target other applications and handle more edge cases.

gantt
    title v0.3 Milestones
    dateFormat YYYY-MM
    section Core
        Config file support (YAML/TOML)  :a1, 2026-04, 3w
        Per-run timeout override         :a2, after a1, 1w
    section Browser
        Firefox / WebKit support         :b1, 2026-04, 4w
        Screenshot diff utility          :b2, after b1, 2w
    section CLI
        --dry-run flag (print goal only) :c1, 2026-05, 1w
        Shell completions (bash/zsh/fish):c2, after c1, 1w
    section Testing
        CLI unit tests (click.testing)   :d1, 2026-04, 2w
        Snapshot regression tests        :d2, after d1, 3w
  • [ ] Config file support — load tot-agent.toml or tot-agent.yaml from the project root so you don't have to pass all options via environment variables.
  • [ ] Cross-browser support — BrowserManager option to launch Firefox or WebKit.
  • [ ] Screenshot diff utility — compare screenshots between runs and highlight changes.
  • [ ] CLI unit tests — use click.testing.CliRunner to test all CLI commands.
  • [ ] --dry-run flag — print the generated goal string without executing.
  • [ ] Shell completions — tot-agent --install-completion bash|zsh|fish.

v0.4 — Parallel execution (2026-Q3)

Theme: Speed up multi-user simulations by running users concurrently.

gantt
    title v0.4 Milestones
    dateFormat YYYY-MM
    section Concurrency
        Async multi-user runner          :a1, 2026-07, 4w
        Rate-limit / semaphore support   :a2, after a1, 2w
    section Reporting
        HTML test report                 :b1, 2026-08, 3w
        JUnit XML output                 :b2, after b1, 1w
    section Observability
        Structured JSON log output       :c1, 2026-07, 2w
        OpenTelemetry tracing            :c2, after c1, 4w
  • [ ] Parallel user simulation — run multiple BrowserAgent instances concurrently with an asyncio.Semaphore to cap API concurrency.
  • [ ] HTML test report — generate a self-contained HTML report with screenshots for each step.
  • [ ] JUnit XML output — CI-friendly test result format for GitHub Actions / Jenkins.
  • [ ] Structured JSON logging — machine-readable log format for log aggregation systems.
  • [ ] OpenTelemetry tracing — distributed tracing for multi-step agent runs.

v0.5 — Platform generalisation (2026-Q4)

Theme: Make it trivial to target any web application without changing Python code.

  • [ ] Goal DSL — a simple YAML/JSON format for defining test scenarios without writing Python.
  • [ ] Plugin system — CoverSource and AgentObserver discoverable via entry points.
  • [ ] Scenario library — shared repository of community-contributed goal templates.
  • [ ] Authentication strategies — OAuth2, MFA simulation, cookie injection.
  • [ ] API mode — expose the agent as a local HTTP service for IDE / CI tool integration.

v1.0 — Stable release (2027-Q1)

Theme: Production-ready, thoroughly documented, actively maintained.

  • [ ] Stable public API with semantic versioning
  • [ ] 90%+ unit test coverage
  • [ ] Comprehensive cookbook / examples directory
  • [ ] Performance benchmarks and regression gates
  • [ ] Security audit: credential handling, browser sandbox configuration
  • [ ] Long-term support commitment

Ideas backlog (unscheduled)

Idea Description
VS Code extension Run goals directly from the editor
Replay mode Record a run and replay it deterministically
Assertion DSL Declarative assertions (e.g. "page must contain 'Success'")
GPT-4o support Optional OpenAI backend via a provider abstraction
Self-healing selectors Automatically update failing selectors using vision
Mobile emulation Playwright device emulation for mobile viewport testing

Contributing

See the GitHub repository to open issues, propose features, or submit pull requests. All roadmap items are tracked as GitHub issues labelled by milestone.