The art-route layer of the Forge skills. route_media.py is the one command that generates an image or an image-to-video clip - through the paid API of any provider whose key is configured (OpenAI, Google Gemini, xAI, BytePlus ModelArk, fal.ai; the configured key is the owner's consent), else through the user's own signed-in Codex or Grok CLI - and prints one JSON line with the route, artifact, sha256 and cost estimate; with no route it answers no-route (codeart2d is then the last resort). Also the capability check (forge_doctor), API key configuration, the spend ledger, resume and Codex image adoption. Use when a Forge skill sends a generation here, for the capability check, or when the user asks about keys, routes or spend. Not for planning, processing or QA of sprites, video frames or maps (generate2dsprite, video2dsprite, generate2dmap) or for code-drawn art (codeart2d).
90%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Medium
Suggest reviewing before use
The route layer. It makes raw media; the owning skill plans the asset before and checks it after.
forge_doctor.py).providers.order goes first). Images: OpenAI (gpt-image-2.5-sunburst), Google Gemini (gemini-3.1-flash-image; --tier hero takes gemini-3-pro-image for hero masters), xAI (grok-imagine-image-2.0), BytePlus ModelArk (Seedream 5.0), fal.ai (reference edits). Video: xAI (grok-imagine-video-1.5; --tier draft takes 1.5 Lite), BytePlus ModelArk (Seedance 2.0), fal.ai (Kling v3, Veo 3.1, Luma, MiniMax, Wan, Vidu, LTX). A configured key is the owner's consent: no per-call question.image_gen, Grok's native tools), otherwise the user's signed-in CLI: Codex (local CLI) image_gen with references attached, then Grok (local CLI) one-shot image or edit; video Grok (local CLI) in ACP mode. An installed CLI is used as is; its first successful run records the VERIFIED proof.route_media.py prints no-route (exit 3).Always name the route that ran ("OpenAI API", "Google Gemini API", "xAI API", "BytePlus ModelArk API", "fal.ai API", "Codex (local CLI)", "Grok (local CLI)"). Details: route-media.md.
Run python "<skill-dir>/scripts/forge_doctor.py" --host-tools <tools> --save <output>/doctor.json, where <tools> lists the media tools in your own tool list (image_gen, image_edit, image_to_video) or none. Its ROUTES block shows which API keys are configured (yes or no, never a key), the local readiness of each CLI and the resolved order (routeOrder). If encoding.stdout fails, set PYTHONUTF8=1.
route_media.py)| Need | Command |
|---|---|
| Which route would run | python "<skill-dir>/scripts/route_media.py" resolve --kind image (or video; --references N) |
| One image, still or edit | python "<skill-dir>/scripts/route_media.py" image --prompt-file <txt> --reference <png> --size 1024x1024 --out-dir <new> (--reference repeatable, in the order the prompt names them) |
| Animate an approved still | python "<skill-dir>/scripts/route_media.py" video --prompt-file <txt> --reference <still.png> --last-frame <still.png> --duration 6 --resolution 720p --out-dir <new> |
| A route or model of the user's choice | add --route api, local, openai, gemini, xai, byteplus, fal, fal:<endpoint-id>, codex-cli, grok-cli or grok-acp; --model <id>, --tier draft or hero, --provider-option [PROVIDER:]KEY=VALUE |
| The plan and estimate first | add --dry-run (no key is used, nothing is sent or written) |
Success prints one ASCII JSON line and exits 0: {"status":"ok","route":"api:openai","artifact":"<out-dir>/generated.png","sha256":...,"estimateUsd":...} plus model, job and, for video, lastFrameUsed, durationRequested and durationUsed (a length is snapped to the nearest one the model renders; Grok (local CLI) renders 6 or 10 s and never pins a last frame) and crop when the frames were padded to 16:9 for a model without 1:1. Every clip is published without audio. No route prints {"status":"no-route","fallback":"codeart2d"} and exits 3. A failure prints one error: <route>: <code>: <message> line and exits 1. A route whose account cannot serve the request (no credit, no access, not signed in) passes it to the next route; a refusal such as moderation stops. Write prompts yourself into UTF-8 files; local routes take the size and every instruction from the prompt.
Each provider reads only its own key: OPENAI_API_KEY, GOOGLE_API_KEY or GEMINI_API_KEY (GOOGLE_API_KEY wins), XAI_API_KEY, ARK_API_KEY (BytePlus ModelArk) and FAL_KEY (fal.ai), from the environment or from the user config file, never from a project: %APPDATA%\agent-sprite-forge\config.json on Windows, ~/.config/agent-sprite-forge/config.json elsewhere ($XDG_CONFIG_HOME when set), as {"OPENAI_API_KEY": "...", "FAL_KEY": "..."} with optional "models" and "providers": {"order": [...]}. Never print, log, commit or paste a key, and never put one in a prompt. A Codex or Grok sign-in is never an API key.
Every call, paid or local, is a line in <project>/.forge/ledger.jsonl with its estimate (estimateUsd is null when no verified price exists; say "price unknown"). There is no cap by default; --budget-usd, --max-calls, FORGE_MAX_PAID_REQUESTS and FORGE_SESSION_IMAGES / FORGE_SESSION_VIDEOS set one when the user asks. python "<skill-dir>/scripts/media_ledger.py" summary shows spend; media_ledger.py settle <id> --status ... closes an unknown outcome.
| Need | Command |
|---|---|
One API request by hand (dry run until --execute) | generate_media.py image or video with --provider, --model, --prompt-file, --out-dir (api-usage.md) |
| Poll or download a video or fal.ai job without paying again | generate_media.py resume --job <out>/job.json |
| One local CLI run by hand | cli_media.py image --route codex-cli --prompt-file <txt> --output-dir <new> --execute (cli-routes.md) |
| A local run timed out or was interrupted | cli_media.py resume --run <id>, then --adopt (never reruns the CLI) |
| Codex made an image that is not in the project | cli_media.py adopt --codex-thread <thread-id> --output-dir <new> |
| Many jobs | generate_media.py batch or cli_media.py batch |
| Verify a local route ahead of time (optional) | forge_doctor.py --verify-route <route> --execute (one quota call) |
image_gen is your own tool: with no configured key use it and adopt the result (cli_media.py adopt); scripts reach Codex (local CLI) through route_media.py. This skill is explicit-only in Codex.route_media.py takes the API or the local CLIs. <skill-dir> is ${CLAUDE_SKILL_DIR}; look at results with Read.grok executable is the "Grok (local CLI)" route for other hosts. Agent setups: agent-profiles/README.md.Run each tool as one line from the user's project root: python "<skill-dir>/scripts/<tool>.py" .... Outputs go to a new --out-dir inside the project. Results are raw media: hand them to the owning skill for identity, size, alpha, motion and seam checks. Keep prompts, input hashes, job ids and result hashes; report the model requested and the model returned separately. Providers: provider-survey.md (last verified 2026-10-06, with the shutdown dates); capabilities: references/capabilities.json (python "<skill-dir>/scripts/media_providers.py" list); prices: references/prices.json; data contracts: references/schemas/media.schema.json.
0x0funky/agent-sprite-forge
5324342
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.