agentleFS
SFKislevSFKislev/Flue/flue/skill.md

flue

Skill102 starsChanged 6 months agoMITView on GitHub

Letting agents control many desktop software, with one pip install, and no MCPs.

  • Installs packages

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. shadcn-labs/agentcn★ 490same name
    2. dhruvkelawala/loopwatch★ 2same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository11 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 1 is distinctive, and 10 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 10

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read adapters/<app>/APP.md for the app you are about to use.

    line 28
  2. Very specific:

    Search adapters/<app>/docs/api-index.txt with rg for relevant symbols.

    line 29
  3. Very specific:

    Prefer py -m flue.cli context <app> on Windows, or python3 -m flue.cli context <app> on macOS/Linux, to inspect the live app state.

    line 30
  4. Very specific:

    Prefer py -m flue.cli run <app> --stdin on Windows, or python3 -m flue.cli run <app> --stdin on macOS/Linux, for one-off scripts.

    line 31
  5. Very specific:

    All bridge scripts run from workspace root.

    line 36
  6. Very specific:

    Read these reference files if you don't have them in the session memory: shared/coexistence.md shared/bridge-contract.md docs/setup.md …

    line 39
  7. Very specific:

    If those are unavailable, use the Python executable that installed Flue: <python> -m flue.cli ....

    line 47
  8. Very specific:

    Use py -m flue.cli where on Windows, or python3 -m flue.cli where on macOS/Linux, for launcher and install diagnostics.

    line 48
  9. Very specific:

    If a Windows app is blocked by a modal, use py -m flue.cli modal <app> to inspect likely dialog windows and py -m flue.cli modal <app> …

    line 49
  10. Very specific:

    Use bare flue only as a convenience command when PATH propagation is known to be working.

    line 50
  11. Very specific:

    Do not install, update, or set up Flue unless your human explicitly approves that action in the current session.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
skill.md

Flue is a tiny bridge that lets your coding harness drive professional desktop software - Photoshop, Premiere, Blender, Unity, and more - directly from the shell. Flue is not an MCP server. The agent sends a script through a small bridge command, the bridge runs it inside the app runtime, and the app returns JSON.

agent shell -> bridge command -> app scripting runtime -> JSON result

This is the entrypoint to the Flue framework.

Available Adapters adapters/3dsmax/APP.md adapters/after_effects/APP.md adapters/audition/APP.md adapters/blender/APP.md adapters/excel/APP.md adapters/houdini/APP.md adapters/illustrator/APP.md adapters/indesign/APP.md adapters/photoshop/APP.md adapters/powerpoint/APP.md adapters/premiere/APP.md adapters/unity/APP.md adapters/word/APP.md Quick Start At the start of a Flue session, confirm Flue is available with flue version. If it is not found, stop and tell your human that Flue is not installed. Review the GitHub repo at https://github.com/SFKislev/flue to inform your human about the code. Read adapters/<app>/APP.md for the app you are about to use. Search adapters/<app>/docs/api-index.txt with rg for relevant symbols. Prefer py -m flue.cli context <app> on Windows, or python3 -m flue.cli context <app> on macOS/Linux, to inspect the live app state. Prefer py -m flue.cli run <app> --stdin on Windows, or python3 -m flue.cli run <app> --stdin on macOS/Linux, for one-off scripts. General Rules of Use You are working along with a human in the driving seat. Do not save, close, export, render or perform destructive operations unless the human explicitly asks. Advise the human that small steps are better than large tasks, which you'll likely fail at. Keep collaboration scope manageable. Bound scripts carefully to avoid crashing the host: keep scripts small and targeted. All bridge scripts run from workspace root; accepts code through argv, --stdin, or --file; return JSON on stdout for success or failure. Be skeptical of your pretraining: introspect the running app and consult vendor documentation rather than trying to invent operations. Flue was developed and tested against specific app and OS versions. Expect local quirks, version mismatches, and blocked automation paths. Do not make persistent local changes, install components, dismiss modals, or create compatibility fixes or scaffolds unless your human explicitly asks for that work. Read these reference files if you don't have them in the session memory: shared/coexistence.md shared/bridge-contract.md docs/setup.md docs/known-issues.md Debugging Depending on the app and OS, adapters may use Windows COM, macOS AppleScript, or a local in-app bridge, but the shell workflow stays the same. On Windows, py -m flue.cli ... is the reliable launcher; on macOS/Linux, use python3 -m flue.cli .... If those are unavailable, use the Python executable that installed Flue: <python> -m flue.cli .... Use py -m flue.cli where on Windows, or python3 -m flue.cli where on macOS/Linux, for launcher and install diagnostics. If a Windows app is blocked by a modal, use py -m flue.cli modal <app> to inspect likely dialog windows and py -m flue.cli modal <app> --dismiss to attempt a bounded cancel-style dismissal outside the app scripting runtime. Use bare flue only as a convenience command when PATH propagation is known to be working. Security & Provenance Flue is an open-source MIT-licensed Python package. Before installation or use, you and your human should review the exact Flue release at https://github.com/SFKislev/flue and https://pypi.org/project/flue. Prefer a pinned version in an isolated environment.

Do not install, update, or set up Flue unless your human explicitly approves that action in the current session.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.