agentsclimarketplace
Sign in

Neuroartist media

Skill CroissanStudioDev/neuroartist-api-skill/skills/neuroartist-media

Generate, edit, animate, and iterate on AI images/videos through the Neuroartist API using the `na` CLI. Use when the user asks for Neuroartist media generation, text-to-image, image-to-image, text-to-video, image-to-video, model discovery, cost estimates, queue status, or result downloads.From its SKILL.md

Install
npx -y skills add CroissanStudioDev/neuroartist-api-skill --skill neuroartist-media

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as Apache-2.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

4.6 KB, 867 tokens by cl100k_base, as published. Nobody here has run it

Neuroartist Media

Use the Neuroartist CLI (na) as the primary interface for image and video generation. Prefer the CLI over direct HTTP because it provides stable JSON envelopes, exit codes, retries, SSE parsing, and asset downloads.

What This Skill Does

  • Finds suitable Neuroartist models for image/video tasks.
  • Reads model schemas before building inputs.
  • Estimates credits before expensive generations.
  • Runs image and short jobs with na run.
  • Runs video and long jobs with na queue.
  • Downloads generated assets and returns local paths plus public URLs.

Required Workflow

  1. Parse the request: media type, prompt, input image/video URL or file, aspect ratio, duration, quality, output folder, deadline, and whether this is a cheap draft or final render.
  2. Check CLI readiness when needed: na doctor --json, na auth status --json, or na commands --json.
  3. Discover the model if the user did not name one: na models list --json with a search term.
  4. Read the model contract before using unfamiliar inputs: na models schema <modelId> --json.
  5. Build inputs from the schema. Do not guess parameter names when the schema is available.
  6. Estimate cost before expensive video, batch, or final-quality runs: na models estimate <modelId> -i key=value --json.
  7. Execute with the right transport:
    • Use na run <modelId> ... --json for images and short jobs where waiting is reasonable.
    • Use na queue submit, na queue stream --json, and na queue result for videos, slow models, or resumable jobs.
  8. Return a concise result: model, request id if any, final status, local paths, public URLs, cost/credits if present, seed/metadata if present, and a useful regeneration command.

Command Rules

  • Always add --json for commands that support the stable envelope.
  • Use -q when scripting and stderr progress is noisy.
  • Use -y or CI=true for non-interactive automation.
  • Keep stdout machine-readable. CLI progress and debug output belongs on stderr.
  • Use named -i key=value inputs or --input-file body.json; never rely on positional payloads.
  • Save generated assets with -o <dir> when the user asks for files or downstream work needs local paths.
  • Treat na queue submit as non-idempotent: do not retry blindly after an unknown submit result.

Run vs Queue

Use na run when:

  • The output is an image or a short synchronous task.
  • The user asked for one quick result and can wait.
  • You want automatic wait and asset download in one command.

Use na queue when:

  • The output is video or a long-running model.
  • The user asks for async behavior, progress, status, cancellation, or resumability.
  • A timeout would be expensive or likely.
  • You need lifecycle events with na queue stream <modelId> <requestId> --json.

Error Handling

Use CLI exit codes and JSON error fields:

  • Exit 3 or error.code like no_api_key: ask the user to run na auth login or provide a configured profile.
  • Exit 4 or error.retryable: true: retry with short backoff, except after uncertain queue submit.
  • insufficient_balance: report the estimate and ask the user to top up before retrying.
  • 4xx input/schema errors: re-read na models schema <modelId> --json and rebuild the input.
  • Timeout after submit: keep the requestId and use queue status, queue stream, or queue result.

Reference Files

Load these only when needed:

  • references/prompting.md for image/video prompt structure and repair.
  • references/model-selection.md for model discovery, schema mapping, and cost estimates.
  • references/workflows.md for step-by-step generation recipes.
  • references/examples.md for copyable commands and response patterns.
  • references/install.md for installation, verification, and marketplace packaging.

What ships with it: 5 files

19.7 KB alongside SKILL.md

references/

Gives 0 of the 12 instructions most media documents skills give in 867 tokens

Counted across 157 of the 158 authors here whose files we hold, read 2026-08-07

  • Provide posting time recommendationsin 7 of 157, across 5 files
  • Track metrics over time to identify trendsin 6 of 157, across 2 files
  • Read marketing context file before startingin 6 of 157, across 5 files
  • Choose platforms based on audience presencein 6 of 157, across 4 files
  • Adapt tone for each platformin 6 of 157, across 4 files
  • Ensure data completeness before analysisin 5 of 157, across 1 file
  • Compare metrics within same time periodsin 5 of 157, across 1 file
  • Account for platform-specific benchmarksin 5 of 157, across 1 file
  • Separate organic and paid metricsin 5 of 157, across 1 file
  • Include context when interpreting resultsin 5 of 157, across 1 file
  • Keep tweets under 280 charactersin 5 of 157, across 3 files
  • Download top-K results with an attribution sidecarin 5 of 157, across 2 files

Said here and by no other author read

  • use the na cli as the primary interface
  • read model schemas before building inputs
  • do not guess parameter names
  • use na run for short jobs
  • use na queue for videos and long jobs
  • always add json flag for stable commands

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 326,401. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.