Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Comprehensive debugging toolkit for Gamma integration issues. Use when you need detailed diagnostics, request tracing, or systematic debugging of Gamma API problems. Trigger with phrases like "gamma debug bundle", "gamma diagnostics", "gamma trace", "gamma inspect", "gamma detailed logs".
.claude/skills/jeremylongshore-gamma-debug-bundle/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 30% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 63% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 133% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 97% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 95% | 0% |
!node --version 2>/dev/null || echo 'N/A' !python3 --version 2>/dev/null || echo 'N/A' !uname -a
Debugging toolkit for Gamma API integration issues. Includes connectivity tests, request tracing, diagnostic scripts, and support ticket templates. Gamma uses a REST API at https://public-api.gamma.app/v1.0/ with X-API-KEY header authentication.
curl and jq availablebash#!/bin/bash set -euo pipefail echo "=== Gamma API Diagnostic ===" # 1. Check API key if [ -z "${GAMMA_API_KEY:-}" ]; then echo "FAIL: GAMMA_API_KEY not set"; exit 1 fi echo "OK: API key set (${#GAMMA_API_KEY} chars, prefix: ${GAMMA_API_KEY:0:4}...)" # 2. Test authentication via /themes endpoint STATUS=$(curl -s -o /dev/null -w "%{http_code}" \ -H "X-API-KEY: $GAMMA_API_KEY" \ "https://public-api.gamma.app/v1.0/themes") if [ "$STATUS" = "200" ]; then echo "OK: Authentication successful" elif [ "$STATUS" = "401" ]; then echo "FAIL: Invalid API key (401)" elif [ "$STATUS" = "403" ]; then echo "FAIL: Account not on Pro+ plan (403)" else echo "WARN: Unexpected status $STATUS" fi # 3. Check latency echo "--- Latency (3 samples) ---" for i in 1 2 3; do LATENCY=$(curl -s -o /dev/null -w "%{time_total}" \ -H "X-API-KEY: $GAMMA_API_KEY" \ "https://public-api.gamma.app/v1.0/themes") echo " Request $i: ${LATENCY}s" done # 4. List themes (verify API access) echo "--- Workspace Themes ---" curl -s -H "X-API-KEY: $GAMMA_API_KEY" \ "https://public-api.gamma.app/v1.0/themes" | jq '.[].name' 2>/dev/null || echo "Could not parse themes"
typescript// scripts/gamma-diagnose.ts const BASE = "https://public-api.gamma.app/v1.0"; const headers = { "X-API-KEY": process.env.GAMMA_API_KEY!, "Content-Type": "application/json", }; interface TestResult { test: string; status: "PASS" | "FAIL"; detail: string; latencyMs?: number; } async function diagnose(): Promise<TestResult[]> { const results: TestResult[] = []; // Test 1: API key set if (!process.env.GAMMA_API_KEY) { results.push({ test: "API Key", status: "FAIL", detail: "GAMMA_API_KEY not set" }); return results; } results.push({ test: "API Key", status: "PASS", detail: `Set (${process.env.GAMMA_API_KEY.length} chars)`, }); // Test 2: Authentication try { const start = Date.now(); const res = await fetch(`${BASE}/themes`, { headers }); const latency = Date.now() - start; if (res.ok) { const themes = await res.json(); results.push({ test: "Authentication", status: "PASS", detail: `OK, ${themes.length} themes available`, latencyMs: latency, }); } else { results.push({ test: "Authentication", status: "FAIL", detail: `HTTP ${res.status}: ${await res.text()}`, }); } } catch (err: any) { results.push({ test: "Authentication", status: "FAIL", detail: err.message }); } // Test 3: Folders endpoint try { const start = Date.now(); const res = await fetch(`${BASE}/folders`, { headers }); results.push({ test: "Folders API", status: res.ok ? "PASS" : "FAIL", detail: res.ok ? `${(await res.json()).length} folders` : `HTTP ${res.status}`, latencyMs: Date.now() - start, }); } catch (err: any) { results.push({ test: "Folders API", status: "FAIL", detail: err.message }); } // Test 4: Generation (dry-run style — creates a minimal generation) try { const start = Date.now(); const res = await fetch(`${BASE}/generations`, { method: "POST", headers, body: JSON.stringify({ content: "Diagnostic test: one card about testing", outputFormat: "presentation", }), }); if (res.ok) { const { generationId } = await res.json(); results.push({ test: "Generation API", status: "PASS", detail: `Started: ${generationId}`, latencyMs: Date.now() - start, }); // Note: this consumes credits — skip in automated tests } else { results.push({ test: "Generation API", status: "FAIL", detail: `HTTP ${res.status}: ${await res.text()}`, }); } } catch (err: any) { results.push({ test: "Generation API", status: "FAIL", detail: err.message }); } // Print report console.log("\n=== Gamma Diagnostic Report ==="); for (const r of results) { const latency = r.latencyMs ? ` (${r.latencyMs}ms)` : ""; console.log(` [${r.status}] ${r.test}: ${r.detail}${latency}`); } const failures = results.filter((r) => r.status === "FAIL").length; console.log(`\n${results.length} tests, ${failures} failures\n`); return results; } diagnose();
Run: npx tsx scripts/gamma-diagnose.ts
typescript// src/gamma/debug-client.ts export function createDebugGammaClient(apiKey: string) { const base = "https://public-api.gamma.app/v1.0"; const headers = { "X-API-KEY": apiKey, "Content-Type": "application/json" }; async function debugRequest(method: string, path: string, body?: unknown) { const start = Date.now(); const url = `${base}${path}`; console.log(`[GAMMA] ${method} ${path}`, body ? JSON.stringify(body).slice(0, 200) : ""); const res = await fetch(url, { method, headers, body: body ? JSON.stringify(body) : undefined, }); const duration = Date.now() - start; const responseText = await res.text(); const status = res.ok ? "OK" : "ERROR"; console.log(`[GAMMA] ${status} ${res.status} in ${duration}ms`); if (!res.ok) { console.log(`[GAMMA] Response: ${responseText.slice(0, 500)}`); } if (!res.ok) throw new Error(`Gamma ${res.status}: ${responseText}`); return JSON.parse(responseText); } return { generate: (body: any) => debugRequest("POST", "/generations", body), poll: (id: string) => debugRequest("GET", `/generations/${id}`), listThemes: () => debugRequest("GET", "/themes"), listFolders: () => debugRequest("GET", "/folders"), }; }
markdown## Environment - Node.js: [version] - OS: [os] - API version: v1.0 ## Issue Description [What you expected vs what happened] ## API Request - Endpoint: POST /v1.0/generations - Status: [HTTP status] - Response: [error body, sanitized] ## Steps to Reproduce 1. [Step 1] 2. [Step 2] ## Diagnostic Output [Paste output from gamma-diagnose.ts] ## Additional Context - Account plan: [Pro/Ultra/Teams/Business] - Credit balance: [approximate]
| Symptom | Likely Cause | Fix | |---------|-------------|-----| | 401 on all requests | Bad API key | Verify key at gamma.app/settings | | 403 Forbidden | Not on Pro+ plan | Upgrade at gamma.app/pricing | | 429 Too Many Requests | Rate limit | Add backoff; contact Gamma support for higher limits | | Generation status: "failed" | Content too complex | Simplify content, reduce card count | | Empty exportUrl | No exportAs in request | Add exportAs: "pdf" to generation | | Timeout on poll | Very complex generation | Increase poll timeout beyond 3 min |
| Error | Cause | Solution | |-------|-------|----------| | GAMMA_API_KEY required | Missing env var | Set GAMMA_API_KEY in .env | | Diagnostic generation fails | No credits | Check credit balance at gamma.app | | Network timeout | Connectivity issue | Check DNS resolution for public-api.gamma.app |
Proceed to gamma-rate-limits for rate limit management.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 25,838 | 18,566 | -28% | 1 | 1 | 0% | 4,190 | 5,445 | +30% | 0 | 0 | — |
case-02 | fail→fail | 25,448 | 18,995 | -25% | 1 | 1 | 0% | 4,529 | 5,499 | +21% | 0 | 0 | — |
case-03 | fail→fail | 29,575 | 22,082 | -25% | 1 | 1 | 0% | 5,334 | 6,104 | +14% | 0 | 0 | — |
case-04 | pass→pass | 15,776 | 22,278 | +41% | 1 | 1 | 0% | 3,275 | 6,176 | +89% | 0 | 0 | — |
case-05 | fail→pass | 26,552 | 26,573 | +0% | 1 | 1 | 0% | 4,079 | 6,630 | +63% | 0 | 0 | — |
case-06 | fail→pass | 11,334 | 16,095 | +42% | 1 | 1 | 0% | 2,026 | 4,730 | +133% | 0 | 0 | — |
case-07 | pass→pass | 9,309 | 5,830 | -37% | 1 | 1 | 0% | 1,500 | 3,524 | +135% | 0 | 0 | — |
case-08 | fail→pass | 12,513 | 14,226 | +14% | 1 | 1 | 0% | 2,029 | 4,005 | +97% | 0 | 0 | — |
case-09 | pass→pass | 5,992 | 2,472 | -59% | 1 | 1 | 0% | 1,068 | 2,934 | +175% | 0 | 0 | — |
case-10 | pass→pass | 18,795 | 13,583 | -28% | 1 | 1 | 0% | 2,122 | 4,705 | +122% | 0 | 0 | — |
case-11 | pass→pass | 15,688 | 14,952 | -5% | 1 | 1 | 0% | 1,905 | 4,422 | +132% | 0 | 0 | — |
case-12 | pass→pass | 4,785 | 8,404 | +76% | 1 | 1 | 0% | 904 | 3,126 | +246% | 0 | 0 | — |
case-13 | fail→pass | 24,600 | 2,509 | -90% | 1 | 1 | 0% | 1,529 | 2,974 | +95% | 0 | 0 | — |
case-14 | fail→pass | 15,134 | 7,972 | -47% | 1 | 1 | 0% | 1,757 | 3,067 | +75% | 0 | 0 | — |
case-15 | pass→pass | 14,863 | 7,636 | -49% | 1 | 1 | 0% | 1,763 | 3,038 | +72% | 0 | 0 | — |
case-16 | pass→pass | 12,188 | 7,578 | -38% | 1 | 1 | 0% | 2,221 | 3,059 | +38% | 0 | 0 | — |
case-17 | fail→fail | 16,932 | 17,241 | +2% | 1 | 1 | 0% | 2,297 | 4,827 | +110% | 0 | 0 | — |
case-18 | fail→pass | 7,109 | 1,728 | -76% | 1 | 1 | 0% | 1,413 | 2,838 | +101% | 0 | 0 | — |
case-19 | pass→fail | 4,807 | 7,270 | +51% | 1 | 1 | 0% | 973 | 2,986 | +207% | 0 | 0 | — |
case-20 | fail→pass | 8,735 | 4,131 | -53% | 1 | 1 | 0% | 1,871 | 3,561 | +90% | 0 | 0 | — |
case-21 | pass→pass | 5,549 | 8,006 | +44% | 1 | 1 | 0% | 992 | 3,076 | +210% | 0 | 0 | — |
case-22 | fail→pass | 16,760 | 6,275 | -63% | 1 | 1 | 0% | 3,282 | 2,742 | -16% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.