FlowLens

Future Work

The next product, reliability, and release steps after the current implementation pass.

Release

Signed public release channel

The app already has Electron package targets. The next release step is macOS Developer ID signing and notarization, Windows code signing, and published GitHub Release artifacts for updates.

Capture

Active-window capture

Improve precision by targeting the active application window while still excluding the overlay reliably.

Providers

Provider adapter registry

Keep the OpenAI-compatible adapter, then formalize the current OpenCode Go and Gemini presets into a clearer registry with provider-specific validation copy.

Product

More focused modes

Extend beyond Prompt Doctor, Error Explainer, and Writing Improver only where a mode has a distinct output contract.

Conversation

Richer follow-up loops

Support more context recovery without turning the product into a heavyweight chat surface.

Integrations

Editor and paste targets

Connect copyable output to editor actions, paste destinations, and safer apply flows.

Privacy

Capture review and redaction

Add optional screenshot confirmation, region selection, and redaction helpers before provider submission.

Reliability

Runtime health surface

Improve retry/backoff, auth error labeling, provider-specific errors, crash recovery, and user-facing status.

The likely next frontier

The next frontier is not "make FlowLens do everything." It is "make the same tight interaction loop safer, easier to install, and more precise."

That means:

  • signed installers and clean update delivery
  • better capture targeting
  • clearer provider presets and compatibility diagnostics
  • stronger privacy controls
  • richer diagnostics and recovery
  • integrations that keep output actionable without becoming autonomous control

If the current build proves the interaction model, future work should deepen that model before widening it.

On this page