Experimental Tools

Experimental Tools – Roadmap

Experimental Tools are a collection of small utilities, nodes and prototypes that I use to solve day-to-day problems in ComfyUI, 3D workflows and AI-assisted production. They’re often rough around the edges, opinionated, and designed for real work rather than general release.

This page gives an overview of what I’m experimenting with, what’s currently being tested, and what might turn into fully supported tools or plugins later on.

🧪 Current Experiments

These tools are actively being used in real projects and are under constant revision. Expect breaking changes, renamed parameters and the odd edge case.

1. Batch Filename & Shot Helper

A set of nodes for managing filenames, shot codes and output folders. Built to keep multi-shot and multi-sequence jobs organised without relying on manual renaming.

  • Auto-constructs paths based on project / shot / version.
  • Keeps output structures consistent across different graphs.
  • Reduces “where did that render go?” moments.

2. Render Notes & Metadata Stamp

Utility nodes for embedding simple metadata into output filenames or sidecar JSON – things like prompt fragments, sampler choices or version notes.

  • Attach a quick note to each render batch.
  • Optional sidecar JSON for pipeline tracking.
  • Useful when iterating on looks with clients or teams.

3. Graph Snapshot & Diff Helper

A lightweight approach to snapshotting “known good” graph settings and comparing them against the current state.

  • Capture a baseline configuration before a big tweak session.
  • Highlight changed sliders / toggles to see what drifted.
  • Early tests for eventual version-control style tools.

4. Utility Prompts & Text Blocks

A small library of reusable text blocks for prompts, tags and style descriptors, aimed at keeping phrasing consistent across graphs.

  • Reusable style fragments for look consistency.
  • Simple “include / exclude” toggles for prompt flavour.
  • Ideal for pipelines with multiple operators touching the same job.

📍 Roadmap

The roadmap is deliberately loose – tools move between stages as they prove themselves in production or reveal new problems to solve.

▶ Now — Actively Prototyping

  • Curve Preset Manager – shared curve presets across multiple nodes and graphs.
  • Batch Render Helper – higher-level wrapper around common batch patterns.
  • Render Checklist – simple pre-flight checks to avoid obvious mistakes.

⏭ Next — Candidates for Public Release

  • Shot & Sequence Toolkit – tools for handling multi-shot jobs with consistent naming.
  • Prompt Library Node – centralised prompt blocks with category tags and quick search.
  • Lightweight Render Logger – writes basic run info to disk for tracking and debugging.

🔭 Later — Ideas on the Back Burner

  • Graph Profiler – basic timing / performance info for heavy pipelines.
  • Shot Review Exporter – auto-pack frames, notes and metadata for client review.
  • ComfyUI & 3D Bridge Utilities – helper nodes to tie 3D scene data into AI passes.

❓ Maybe — Experiments & Wild Cards

  • Node templates for “good defaults” on common tasks.
  • Higher-level UI wrappers for non-technical artists.
  • Deeper integration with external pipeline tools if demand appears.

🧵 How Experimental Tools Are Used

These utilities live in the space between “quick hack” and “shippable product”. They often start life as a one-off script to solve a problem on a job, then gradually harden into something reusable.

  • Everything is subject to change. Names, inputs and behaviour may shift.
  • Tools are tested on real client or personal projects before being considered stable.
  • Some experiments will never ship – they’re stepping stones for better ideas.

📫 Early Access & Feedback

If you’re a studio, developer or artist interested in testing in-development tools, I occasionally share early builds with a small group of people who don’t mind bugs and rough edges.

If that sounds like you, feel free to get in touch via the contact page and mention that you’re interested in Experimental Tools. Let me know what kind of work you’re doing and which parts of the pipeline you’d like to improve – it helps me prioritise what to stabilise next.