agentleFS
badgidsbadgids/Story-Film-Skills/skills/comfyui-cli/SKILL.md

comfyui-cli

Skill4 starsChanged 3 months agoApache-2.0View on GitHub

Use the official comfy-cli safely for ComfyUI workspace discovery, setup, launch and stop, structured JSON automation, workflow execution including UI-format conversion, jobs, uploads, downloads, nodes, models, templates, memory, updates, and approved dependency changes.

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 6 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask the installed CLI for help when command syntax is uncertain.

    line 19
  2. Very specific:

    Prefer --json for machine-readable automation.

    line 20
  3. Distinctive:

    Use background launch for an agent-controlled server lifecycle.

    line 22
  4. Distinctive:

    Use comfy-cli's run path when a UI-format workflow needs its supported conversion.

    line 24
  5. Distinctive:

    Keep the prompt ID and use jobs/download for long runs.

    line 25
  6. Very specific:

    Do not add --where cloud as an automatic fallback.

    line 44
  7. Distinctive:

    Use the bridge's read-only model and package discovery before changing the environment:

    line 48
  8. Distinctive:

    After a custom-node install, update, reinstall, repair, or dependency installation, restart ComfyUI and run live workflow validation again.

    line 58

≈ “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

comfy-cli

Read

  • ../../references/COMFYUI_CLI.md
  • ../../references/COMFYUI_SECURITY.md

Procedure

  1. Detect comfy from COMFY_BIN or PATH. Do not guess a personal binary path.
  2. Ask the installed CLI for help when command syntax is uncertain.
  3. Prefer --json for machine-readable automation.
  4. Resolve the target workspace with comfy which or the user's explicit workspace option.
  5. Use background launch for an agent-controlled server lifecycle.
  6. Probe ComfyUI after launch.
  7. Use comfy-cli's run path when a UI-format workflow needs its supported conversion.
  8. Keep the prompt ID and use jobs/download for long runs.

Bundled bridge:

python scripts/comfyui_cli_bridge.py info
python scripts/comfyui_cli_bridge.py launch
python scripts/comfyui_cli_bridge.py stop
python scripts/comfyui_cli_bridge.py run --workflow WORKFLOW.json

The bridge is intentionally narrow. For commands not wrapped there, inspect the installed CLI help and call comfy directly.

Mutating dependencies

Custom node installs, broad updates, version switching, and model downloads change the execution environment. Require user intent, then restart and revalidate as appropriate.

Cloud

Do not add --where cloud as an automatic fallback. Cloud and partner operations can spend credits.

Live dependency operations

Use the bridge's read-only model and package discovery before changing the environment:

python scripts/comfyui_cli_bridge.py models-search --text MODEL
python scripts/comfyui_cli_bridge.py node-deps PACK
python scripts/comfyui_cli_bridge.py deps-in-workflow --workflow WORKFLOW.json --out DEPS.json

For an approved custom-node or model change, use the guarded bridge command. Never add --confirm merely because a command refused to proceed. The user must already have approved that exact environment change.

After a custom-node install, update, reinstall, repair, or dependency installation, restart ComfyUI and run live workflow validation again.

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.