Agent skill · software engineering · jeremylongshore
canva-performance-tuning
Optimize Canva Connect latency and throughput from measured application evidence. Use when tuning metadata caches, pagination, connection reuse, async-job polling, or endpoint concurrency under explicit freshness and safety budgets. Trigger with: "speed up Canva", "tune Canva polling", "optimize Canva performance".
Why this skill is useful
Adds executable scripts for optimizing Canva API performance with caching and pagination strategies that aren't in public docs.
What it needs
Requires ioredis, lru-cache, p-queue, redis installed locally. Requires canva account access. About 2k tokens when loaded. Last updated 2026-08-07. 2,604 stars on the source repository.
What this skill does
Canva Measured Performance Tuning Overview Change one controlled variable at a time and compare logical-operation outcomes, not just raw request latency. Never cache credentials or assume fixed Canva URL lifetimes and performance guarantees. Prerequisites Baseline window, normalized endpoint, workload, and local SLO Current OpenAPI/response contract and endpoint rate metadata Cache data class, freshness/invalidation policy, feature flag, and rollback Instructions Step 1: Establish a baseline Use Read and Grep to measure logical operations, provider calls, latency distribution, job completion, retries, cache hits, errors, and queue depth with synthetic or approved data. Step 2: Identify the bottleneck Separate network/connection latency, unnecessary fields, pagination, duplicate reads, write retries, polling cadence, token locks, worker capacity, and local storage. Step 3: Design one experiment Use Write or Edit to change one cache, pagination, connection, queue, or poll control behind a feature flag with explicit success and rollback thresholds. Step 4: Protect caches Cache only approved metadata, key by authorization boundary, encrypt where policy requires, and invalidate on writes, consent/ownership changes, or freshness expiry. Step 5: Tune asynchronous polling Persist job ID, begin with a short local interval, apply bounded exponential backoff, and stop at the application budget while reconciliation continues asynchronously. Step 6: Validate and roll out Compare correctness, freshness, duplicate prevention, resource use, and latency. Roll back on stale authorization, missed completion, increased errors, or throttling. Authentication Canva Connect calls use Bearer access tokens obtained by a backend through OAuth 2.0 Authorization Code with SHA-256 PKCE. Request explicit least-privilege scopes, keep client secrets and tokens out of browser-visible state, and serialize refresh so the replacement single-use refresh token is stored atomically. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills jeremylongshore/canva-performance-tuning