Agent skill · software engineering · jeremylongshore
anima-rate-limits
Implement rate limiting for Anima API code generation requests.
Why this skill is useful
Adds executable scripts for rate limiting and batching Anima API requests, which are not covered in public documentation.
What it needs
Requires bottleneck installed locally. Requires anima account access. About 3k tokens when loaded. Last updated 2026-08-07. 2,604 stars on the source repository.
What this skill does
Anima Rate Limits Overview Handle Figma rate limits from the current SDK's structured callback and treat any Anima generation quota as account-specific until the provider or agreement says otherwise. Never encode an invented requests-per-minute or concurrency allowance. Documented Signals Surface Signal Required action --------- -------- ----------------- FigmaRestApi.onRateLimited retryAfter, figmaPlanTier, figmaRateLimitType Retry only inside a bounded elapsed-time budget figmaRateLimitMaxWait 1–180 seconds; default 60 Set explicitly for the request's latency budget Anima generation Returned error/progress state and account contract Stop or queue; do not infer a numeric quota Prerequisites Confirm the account's current generation and concurrency contract before choosing worker concurrency; absence of a published number means fail closed, not guess. Store ANIMATOKEN, FIGMATOKEN, and FIGMAFILEKEY in the runtime secret manager or injected environment, and verify that the token has only the scopes required for the selected file. Never put tokens in source, fixtures, logs, generated receipts, or retry payloads. Define an explicit allowlist of Figma files and node IDs, a bounded batch size, a maximum retry budget, and an approved output directory. Use synthetic or sandbox designs for load tests and confirm that generated output contains no customer data before persisting it. Install the pinned SDK and decide whether a generation is safe to repeat. Persist a redacted request fingerprint and resume only failed node IDs. Instructions Step 1: Configure Structured Figma Rate Handling Step 2: Bind Each Generation to a Maximum Wait Error Handling Let FigmaRestApi own Figma retry decisions. Do not add a second uncoordinated retry loop around the same request. Do not retry 401/403 authentication or permission failures, invalid node/file parameters, or policy rejections. Stop the batch, report the node ID and redacted status, and repair credentials or scope before resuming. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills jeremylongshore/anima-rate-limits--ae66a2