Comfy-agent generate_flux_multi_input_img2img

install

source · Clone the upstream repo

git clone https://github.com/steliosot/comfy-agent

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/steliosot/comfy-agent "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/workflows/img2img_inpaint_outpaint/generate_flux_multi_input_img2img" ~/.claude/skills/steliosot-comfy-agent-generate-flux-multi-input-img2img && rm -rf "$T"

manifest: skills/workflows/img2img_inpaint_outpaint/generate_flux_multi_input_img2img/SKILL.md

source content

generate_flux_multi_input_img2img

Generate one image from 2 or 3 already-uploaded Comfy input images using a FLUX multi-reference img2img pipeline.

Usage

python3 skills/workflows/generate_flux_multi_input_img2img/scripts/run.py --args '{}' --pretty

Pass JSON kwargs for

run(...)

via

--args

python3 skills/workflows/generate_flux_multi_input_img2img/scripts/run.py --args '{"param":"value"}' --pretty

Inputs / Outputs

Inputs follow

skills/workflows/generate_flux_multi_input_img2img/skill.yaml

Output is JSON printed to stdout from
```
run(...)
```
.

Composition

This skill is designed to compose with other single-purpose skills. Use

run_id

and artifact references when chaining.

Routing Metadata

Family:
```
img2img_inpaint_outpaint
```
Input modalities:
```
text_prompt
```
Output modalities:
```
application/json
```
Model families:
```
other
```
Node count:
```
None
```
Complexity score:
```
None
```
Resource profile:
```
unknown
```

Estimated runtime:

depends on runtime parameters and server resources

Max latent resolution hint:
```
None
```
x
```
None
```
Max sampler steps hint:
```
None
```

Detected Models

None detected.

Detected Custom Nodes

None detected.

Runtime Warnings

Hand-authored skill (no embedded workflow.json); metadata is inferred and may be approximate.