# Aavegotchi 3d Renderer > Render Aavegotchi assets by deriving renderer hashes from Goldsky Base core data and calling POST /api/renderer/batch on www.aavegotchi.com. Use when the user gives a tokenId or inventory URL, or when deterministic hash plus image artifacts are required. - Skill: `modbender/aavegotchi-3d-renderer` (Agent Skill, multi-file: 3 files) - Install (CLI): `npx skillmds@latest add modbender/aavegotchi-3d-renderer` - Raw SKILL.md: https://api.skillmd.com/api/skills/modbender/aavegotchi-3d-renderer/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Integrations & APIs - Author: modbender (https://skillmd.com/u/modbender) - Updated: 2026-09-09 - Page: https://skillmd.com/skills/modbender/aavegotchi-3d-renderer --- # Aavegotchi 3D Renderer Render gotchi assets from token data and renderer batch APIs. ## Inputs - Accept either `tokenId` or inventory URL with `id=`. - Target Base by default (`chainId=8453`). ## Outputs - Return derived renderer hash. - Return `/api/renderer/batch` HTTP status. - Save raw batch JSON to disk. - Save `PNG_Full` and `PNG_Headshot` to disk when available. - Return `GLB_3DModel` availability and URL when present. ## Execute 1. Extract `tokenId` from direct input or inventory URL. 2. Query Goldsky Base core subgraph: `https://api.goldsky.com/api/public/project_cmh3flagm0001r4p25foufjtt/subgraphs/aavegotchi-core-base/prod/gn` 3. Derive hash in renderer format: `---------` 4. Kick off render with `POST https://www.aavegotchi.com/api/renderer/batch` using: - `force: true` - `verify: false` - `renderTypes: ["PNG_Full", "PNG_Headshot", "GLB_3DModel"]` 5. Poll `POST /api/renderer/batch` with `verify: true` until `availability.exists=true` for all requested render types or timeout. 6. Download `proxyUrls.PNG_Full` and `proxyUrls.PNG_Headshot` only when corresponding `availability.exists=true`. 7. Return the hash, kickoff + verify responses, poll summary, and saved artifact paths. ## Command Run the bundled script: ```bash node scripts/render-gotchi-bypass.mjs --token-id 6741 ``` Or pass an inventory URL: ```bash node scripts/render-gotchi-bypass.mjs \ --inventory-url "https://www.aavegotchi.com/u/0x.../inventory?itemType=aavegotchis&chainId=8453&id=6741" ``` Use `--out-dir /tmp` to control artifact location (default: `/tmp`). Optional polling controls: ```bash --poll-attempts 18 --poll-interval-ms 10000 ``` ## Return format Always return: 1. `tokenId` 2. `hash` 3. Kickoff status, verify status, and raw JSON paths 4. Poll summary (`pollAttempts`, `pollIntervalMs`, `renderReady`) 5. `PNG_Full` and `PNG_Headshot` output paths (or missing reason) 6. `GLB_3DModel` availability and URL when present ## Troubleshooting - If Goldsky returns no gotchi, verify `tokenId` and Base context. - If batch returns hash-format `400`, verify eye mappings and right/left wearable order (`index4` then `index5`). - If `availability.exists` is `false`, ensure kickoff used `force:true`, then keep polling `verify:true` until timeout. - If endpoint returns `404`, verify production deployment state.