Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Post-setup credential manager. Shows current integration status (configured/missing/expired) and lets you update individual credentials without re-running the full setup wizard. Runs a smoke test after each update.
.claude/skills/davepoon-ops-settings/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -10% | 0% |
!PREFS="${CLAUDE_PLUGIN_DATA_DIR:-$HOME/.claude/plugins/data/ops-ops-marketplace}/preferences.json" cat "$PREFS" 2>/dev/null || echo '{}'
Manage credentials and integration config after initial setup.
--status or empty → show full credential status dashboard<integration-name> → jump directly to updating that integration (e.g. /ops:settings stripe)--status <integration-name> → show status of one integration onlyRead preferences.json. For each known integration, check whether the key exists and is non-empty. Also probe liveness where possible.
Display as a table:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
OPS ► SETTINGS — Integration Status
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Integration Status Last Updated
─────────────────── ──────────── ─────────────
GitHub (gh cli) ✅ active (always active if gh auth status)
Stripe ✅ configured 2026-04-14
RevenueCat ✅ configured 2026-04-14
Telegram ✅ configured 2026-04-13
Slack ⚠️ missing —
Linear ✅ configured 2026-04-11
Sentry ⚠️ missing —
AWS ✅ active (always active if aws sts works)
Shopify ⚠️ missing —
Klaviyo ⚠️ missing —
Meta Ads ⚠️ missing —
GA4 ⚠️ missing —
ElevenLabs ⚠️ missing —
Datadog ⚠️ missing —
New Relic ⚠️ missing —
...
✅ N configured ⚠️ N missing
──────────────────────────────────────────────────────For integrations with a cheap health check, run it to distinguish "configured but expired" from "configured and active":
| Integration | Probe | Active signal | |-------------|-------|---------------| | Stripe | curl -s -o /dev/null -w "%{http_code}" -u "${stripe_key}:" https://api.stripe.com/v1/balance | 200 | | GitHub | gh auth status 2>&1 | "Logged in" | | AWS | aws sts get-caller-identity --output text 2>/dev/null | exits 0 | | Linear | cat "$PREFS" | jq -r .linear_team | non-empty | | Doppler MCP | Check if DOPPLER_TOKEN is set and valid | Token present and MCP server responds |
Show 🔴 expired if probe fails for a previously-configured key.
When a specific integration is selected (via argument or user pick from dashboard):
sk_live_•••••••••••••••• (last 4 chars visible) [Enter new value] [Test current value] [Clear this credential] [Back to dashboard]
AskUserQuestion text inputpreferences.json via jq update:bash tmp=$(mktemp) jq --arg v "$NEW_VALUE" --arg k "$KEY_NAME" '.[$k] = $v' "$PREFS" > "$tmp" && mv "$tmp" "$PREFS"
✅ Stripe key updated — smoke test passed or ⚠️ Key saved but smoke test failed: <reason>| Integration | Smoke test command | |-------------|-------------------| | Stripe | curl -s -u "${new_key}:" https://api.stripe.com/v1/balance \| jq .object → must be "balance" | | RevenueCat | curl -s -H "Authorization: Bearer ${new_key}" "https://api.revenuecat.com/v2/projects" \| jq '.items \| length' → non-zero | | Telegram | node ${CLAUDE_PLUGIN_ROOT}/telegram-server/index.js --health 2>&1 → "healthy" | | Slack | curl -s -H "Authorization: Bearer ${new_token}" https://slack.com/api/auth.test \| jq .ok → true | | Shopify | curl -s -H "X-Shopify-Access-Token: ${new_token}" "https://${store_url}/admin/api/2024-01/shop.json" \| jq .shop.name → non-null | | Klaviyo | curl -s -H "Authorization: Klaviyo-API-Key ${new_key}" https://a.klaviyo.com/api/accounts/ \| jq '.data[0].id' → non-null | | Datadog | curl -s -H "DD-API-KEY: ${new_key}" https://api.datadoghq.com/api/v1/validate \| jq .valid → true | | New Relic | curl -s -H "Api-Key: ${new_key}" https://api.newrelic.com/v2/applications.json \| jq '.applications | length' → numeric | | Doppler MCP | npx -y @dopplerhq/mcp-server --help 2>&1 with DOPPLER_TOKEN set | exits 0 |
| Command | Purpose | |---------|---------| | cat "$PREFS" \| jq 'keys' | List all configured keys | | jq --arg v "$V" --arg k "$K" '.[$k] = $v' "$PREFS" | Update a single key | | gh auth status | Verify GitHub CLI auth | | aws sts get-caller-identity | Verify AWS auth |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,255 | 8,709 | +39% | 1 | 1 | 0% | 1,136 | 1,626 | +43% | 0 | 0 | — |
case-02 | fail→fail | 8,606 | 6,771 | -21% | 1 | 1 | 0% | 1,181 | 1,560 | +32% | 0 | 0 | — |
case-03 | fail→fail | 7,587 | 12,497 | +65% | 1 | 1 | 0% | 1,122 | 1,597 | +42% | 0 | 0 | — |
case-04 | fail→pass | 6,459 | 4,009 | -38% | 1 | 1 | 0% | 1,500 | 2,053 | +37% | 0 | 0 | — |
case-05 | pass→pass | 7,975 | 2,587 | -68% | 1 | 1 | 0% | 1,630 | 1,817 | +11% | 0 | 0 | — |
case-06 | pass→pass | 4,069 | 2,978 | -27% | 1 | 1 | 0% | 791 | 1,891 | +139% | 0 | 0 | — |
case-07 | pass→pass | 4,596 | 2,403 | -48% | 1 | 1 | 0% | 665 | 1,741 | +162% | 0 | 0 | — |
case-08 | pass→pass | 8,191 | 3,282 | -60% | 1 | 1 | 0% | 1,605 | 2,027 | +26% | 0 | 0 | — |
case-09 | pass→pass | 7,327 | 2,141 | -71% | 1 | 1 | 0% | 1,330 | 1,757 | +32% | 0 | 0 | — |
case-10 | pass→pass | 5,412 | 3,474 | -36% | 1 | 1 | 0% | 854 | 1,778 | +108% | 0 | 0 | — |
case-11 | fail→pass | 12,293 | 3,598 | -71% | 1 | 1 | 0% | 1,895 | 2,097 | +11% | 0 | 0 | — |
case-12 | pass→pass | 13,412 | 7,139 | -47% | 1 | 1 | 0% | 2,542 | 2,690 | +6% | 0 | 0 | — |
case-13 | fail→pass | 16,441 | 2,286 | -86% | 1 | 1 | 0% | 2,618 | 1,743 | -33% | 0 | 0 | — |
case-14 | fail→fail | 5,173 | 1,932 | -63% | 1 | 1 | 0% | 792 | 1,624 | +105% | 0 | 0 | — |
case-15 | fail→pass | 21,550 | 3,015 | -86% | 1 | 1 | 0% | 1,909 | 1,878 | -2% | 0 | 0 | — |
case-16 | fail→pass | 13,827 | 2,981 | -78% | 1 | 1 | 0% | 1,936 | 1,735 | -10% | 0 | 0 | — |
case-17 | fail→pass | 14,932 | 3,625 | -76% | 1 | 1 | 0% | 2,237 | 1,915 | -14% | 0 | 0 | — |
case-18 | fail→fail | 9,928 | 3,783 | -62% | 1 | 1 | 0% | 1,711 | 2,049 | +20% | 0 | 0 | — |
case-19 | pass→pass | 12,193 | 3,025 | -75% | 1 | 1 | 0% | 1,585 | 1,922 | +21% | 0 | 0 | — |
case-20 | fail→pass | 7,548 | 4,408 | -42% | 1 | 1 | 0% | 1,288 | 1,565 | +22% | 0 | 0 | — |
case-21 | pass→fail | 6,580 | 4,635 | -30% | 1 | 1 | 0% | 894 | 1,609 | +80% | 0 | 0 | — |
case-22 | pass→pass | 8,072 | 2,966 | -63% | 1 | 1 | 0% | 1,247 | 1,962 | +57% | 0 | 0 | — |
case-23 | pass→pass | 4,217 | 5,263 | +25% | 1 | 1 | 0% | 899 | 2,059 | +129% | 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. 23 cases were attempted, and 19 counted toward the lift figure. The other 4 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +26 percentage points is the difference between those two pass rates over the 19 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.