Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Provision SaaS services + sync creds via Stripe Projects.
.claude/skills/nousresearch-stripe-projects/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 66% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 5% | 0% |
Wraps the Stripe Projects CLI plugin so Hermes can provision SaaS services (Neon, Twilio, Vercel, etc.), generate and sync credentials into the user's .env, and manage billing across providers from one place.
Gated [linux, macos] while the broader payments cluster matures on Windows. The Stripe CLI itself is cross-platform; this gate is a posture for the cluster, not a hard limit.
Trigger phrases:
If the user already has a provider account, this skill can still connect it with stripe projects link <provider>. If the user wants to use an existing provider resource, such as an existing database or Vercel project, check provider support first; many providers currently support provisioning new resources but not importing existing ones.
macOS:
brew install stripe/stripe-cli/stripe
stripe plugin install projectsLinux: follow the platform-specific install at https://docs.stripe.com/stripe-cli/install, then:
stripe plugin install projectsAll commands run through the terminal tool from inside the user's project directory (the CLI writes .env and .projects/vault/vault.json into the CWD).
cd <project-root>
stripe projects initThis creates .projects/vault/vault.json (encrypted credential store) and prepares the project to receive providers.
stripe projects catalogLists every provider Stripe Projects supports — databases, hosting, auth, AI, analytics, messaging, etc.
stripe projects add <provider>/<service>Examples:
stripe projects add neon/postgresstripe projects add twilio/smsstripe projects add runloop/sandboxThe CLI provisions the service in the user's own account with the provider, generates credentials, syncs them into .env, and records the resource in the vault. The user may need to confirm a tier selection or pricing prompt.
stripe projects listShould show the newly added provider and its .env keys.
stripe projects upgrade <provider> # tier change
stripe projects remove <provider> # deprovision
stripe projects rotate <provider> # rotate credentials.env writes are real writes. The CLI appends to whatever .env is in the project root. If the user's .env is gitignored (normal), the keys land safely; if not, this skill could be a credential-leak vector. Always check .gitignore first..projects/vault/vault.json is per-project. Provisioning the same service in two different projects creates two separate resources — and two bills.add/upgrade are real charges; surface them to the user before confirming.stripe projects catalog | grep <name> first instead of failing the add call..env is plaintext. Standard .env hygiene applies — never commit it.remove for high-cost services (managed databases especially).stripe projects --version && stripe projects listExit code 0 inside an initialized project means the plugin is healthy.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 10,765 | 4,906 | -54% | 1 | 1 | 0% | 1,873 | 1,252 | -33% | 0 | 0 | — |
case-02 | fail→fail | 10,276 | 4,456 | -57% | 1 | 1 | 0% | 2,035 | 1,236 | -39% | 0 | 0 | — |
case-03 | fail→fail | 10,184 | 4,029 | -60% | 1 | 1 | 0% | 1,689 | 1,131 | -33% | 0 | 0 | — |
case-04 | fail→fail | 2,574 | 3,210 | +25% | 1 | 1 | 0% | 195 | 1,350 | +592% | 0 | 0 | — |
case-05 | pass→pass | 4,289 | 2,499 | -42% | 1 | 1 | 0% | 670 | 1,208 | +80% | 0 | 0 | — |
case-06 | fail→pass | 4,853 | 1,720 | -65% | 1 | 1 | 0% | 745 | 1,234 | +66% | 0 | 0 | — |
case-07 | fail→pass | 15,472 | 1,857 | -88% | 1 | 1 | 0% | 2,610 | 1,215 | -53% | 0 | 0 | — |
case-08 | fail→pass | 8,699 | 2,885 | -67% | 1 | 1 | 0% | 1,549 | 1,399 | -10% | 0 | 0 | — |
case-09 | fail→pass | 9,108 | 1,727 | -81% | 1 | 1 | 0% | 1,467 | 1,212 | -17% | 0 | 0 | — |
case-10 | fail→pass | 7,315 | 2,243 | -69% | 1 | 1 | 0% | 1,230 | 1,296 | +5% | 0 | 0 | — |
case-11 | pass→pass | 6,772 | 1,484 | -78% | 1 | 1 | 0% | 1,144 | 1,115 | -3% | 0 | 0 | — |
case-12 | fail→pass | 28,308 | 4,076 | -86% | 1 | 1 | 0% | 1,854 | 1,273 | -31% | 0 | 0 | — |
case-13 | fail→pass | 11,457 | 5,536 | -52% | 1 | 1 | 0% | 1,829 | 1,854 | +1% | 0 | 0 | — |
case-14 | fail→pass | 15,108 | 1,320 | -91% | 1 | 1 | 0% | 2,678 | 1,148 | -57% | 0 | 0 | — |
case-15 | fail→pass | 8,101 | 1,178 | -85% | 1 | 1 | 0% | 1,354 | 1,133 | -16% | 0 | 0 | — |
case-16 | fail→fail | 9,620 | 3,784 | -61% | 1 | 1 | 0% | 1,729 | 1,623 | -6% | 0 | 0 | — |
case-17 | pass→pass | 8,529 | 4,158 | -51% | 1 | 1 | 0% | 1,493 | 1,350 | -10% | 0 | 0 | — |
case-18 | fail→pass | 4,402 | 1,629 | -63% | 1 | 1 | 0% | 743 | 1,269 | +71% | 0 | 0 | — |
case-19 | pass→pass | 11,452 | 8,277 | -28% | 1 | 1 | 0% | 2,360 | 2,801 | +19% | 0 | 0 | — |
case-20 | pass→pass | 6,641 | 4,798 | -28% | 1 | 1 | 0% | 1,229 | 1,992 | +62% | 0 | 0 | — |
case-21 | pass→pass | 10,452 | 7,388 | -29% | 1 | 1 | 0% | 2,286 | 2,626 | +15% | 0 | 0 | — |
case-22 | pass→pass | 5,524 | 5,640 | +2% | 1 | 1 | 0% | 1,275 | 2,063 | +62% | 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, and 18 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 +45 percentage points is the difference between those two pass rates over the 18 comparable cases.
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.