Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Wire encrypted credentials to environment variables using the credential: convention. Use when setting up LLM provider keys (ANTHROPIC_API_KEY, OPENAI_API_KEY) or any env var that should come from encrypted credentials.
.claude/skills/growthxai-output-credentials-env-vars/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 47% | 0% |
ANTHROPIC_API_KEY or OPENAI_API_KEY from encrypted credentialsprocess.env variable automatically.env secrets to encrypted credentialscredential: ConventionAny env var whose value starts with credential: is resolved from encrypted credentials at worker startup. The format is:
ENV_VAR_NAME=credential:<dot.path>.envbash# These are resolved automatically from config/credentials.yml.enc ANTHROPIC_API_KEY=credential:anthropic.api_key OPENAI_API_KEY=credential:openai.api_key # Any credential path works MY_SERVICE_TOKEN=credential:my_service.token DATABASE_URL=credential:postgres.url
config/credentials.yml.enc)yamlanthropic: api_key: sk-ant-... # → resolves ANTHROPIC_API_KEY openai: api_key: sk-... # → resolves OPENAI_API_KEY my_service: token: tok_live_... # → resolves MY_SERVICE_TOKEN postgres: url: postgres://... # → resolves DATABASE_URL
.env via dotenv — ANTHROPIC_API_KEY = "credential:anthropic.api_key"@outputai/credentials)runStartupHooks() — resolveCredentialRefs() runsresolveCredentialRefs() scans process.env for credential: prefix valuesANTHROPIC_API_KEY is now "sk-ant-..." in process.env_env Section in Credentials YAMLThe credentials file can also declare the mapping directly in an _env section. New projects scaffold with this pre-configured:
yamlanthropic: api_key: sk-ant-... openai: api_key: sk-... _env: ANTHROPIC_API_KEY: anthropic.api_key OPENAI_API_KEY: openai.api_key
> Note: The _env section is metadata only — it documents the intended mapping but does not drive resolution. Resolution is driven by the credential: values in .env. Keep both in sync.
Real env var values always take precedence. If ANTHROPIC_API_KEY is already set to a non-credential: value (e.g. from the shell or a CI secret), it is never overwritten:
bash# Real value — never touched by resolveCredentialRefs ANTHROPIC_API_KEY=sk-ant-real-override # Placeholder — gets replaced at startup ANTHROPIC_API_KEY=credential:anthropic.api_key
This means you can override any credential ref at deploy time without changing files.
After the first resolution, ANTHROPIC_API_KEY contains the real API key string — it no longer starts with credential:. Subsequent calls to resolveCredentialRefs() are no-ops for that variable.
bashnpx output credentials init npx output credentials edit # Add anthropic.api_key, openai.api_key
.envbash# Replace plaintext secrets with credential references ANTHROPIC_API_KEY=credential:anthropic.api_key OPENAI_API_KEY=credential:openai.api_key
Start the worker and look for the log line:
Startup hooks resolved env vars {"vars":["ANTHROPIC_API_KEY","OPENAI_API_KEY"]}If the log line appears, credentials are wired correctly.
If you need to call resolveCredentialRefs() outside of a worker context:
typescriptimport { resolveCredentialRefs } from '@outputai/credentials'; // Returns array of env var names that were resolved const resolved = resolveCredentialRefs(); console.log('Resolved:', resolved); // → ["ANTHROPIC_API_KEY", "OPENAI_API_KEY"]
config/credentials.yml.enc contains the target credential paths.env uses credential:<path> values for the relevant env varsStartup hooks resolved env varsANTHROPIC_API_KEY is set correctly)output-credentials-init — Create the encrypted credentials fileoutput-credentials-edit — Add/update credential valuesoutput-dev-credentials — Full credentials system reference| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-10 | fail→pass | 18,954 | 9,529 | -50% | 1 | 1 | 0% | 2,806 | 2,795 | -0% | 0 | 0 | — |
case-01 | fail→pass | 16,678 | 5,849 | -65% | 1 | 1 | 0% | 2,675 | 2,195 | -18% | 0 | 0 | — |
case-03 | fail→pass | 19,071 | 9,067 | -52% | 1 | 1 | 0% | 3,102 | 2,847 | -8% | 0 | 0 | — |
case-02 | fail→pass | 18,658 | 6,232 | -67% | 1 | 1 | 0% | 3,332 | 2,223 | -33% | 0 | 0 | — |
case-04 | fail→fail | 15,724 | 7,850 | -50% | 1 | 1 | 0% | 2,659 | 2,115 | -20% | 0 | 0 | — |
case-05 | fail→pass | 7,919 | 3,974 | -50% | 1 | 1 | 0% | 1,258 | 1,843 | +47% | 0 | 0 | — |
case-06 | pass→pass | 15,237 | 9,534 | -37% | 1 | 1 | 0% | 2,462 | 2,857 | +16% | 0 | 0 | — |
case-07 | fail→fail | 11,919 | 3,424 | -71% | 1 | 1 | 0% | 2,115 | 1,815 | -14% | 0 | 0 | — |
case-08 | fail→pass | 15,674 | 4,494 | -71% | 1 | 1 | 0% | 2,546 | 1,895 | -26% | 0 | 0 | — |
case-09 | pass→pass | 16,690 | 6,417 | -62% | 1 | 1 | 0% | 2,640 | 2,302 | -13% | 0 | 0 | — |
case-11 | fail→pass | 14,266 | 4,522 | -68% | 1 | 1 | 0% | 2,225 | 2,018 | -9% | 0 | 0 | — |
case-12 | pass→pass | 5,684 | 2,477 | -56% | 1 | 1 | 0% | 1,014 | 1,618 | +60% | 0 | 0 | — |
case-13 | fail→pass | 11,996 | 2,936 | -76% | 1 | 1 | 0% | 1,912 | 1,636 | -14% | 0 | 0 | — |
case-14 | pass→pass | 11,737 | 2,664 | -77% | 1 | 1 | 0% | 1,760 | 1,587 | -10% | 0 | 0 | — |
case-15 | fail→pass | 14,639 | 5,102 | -65% | 1 | 1 | 0% | 2,455 | 2,116 | -14% | 0 | 0 | — |
case-16 | fail→pass | 11,928 | 6,943 | -42% | 1 | 1 | 0% | 1,822 | 2,392 | +31% | 0 | 0 | — |
case-17 | fail→fail | 16,156 | 3,678 | -77% | 1 | 1 | 0% | 2,819 | 1,821 | -35% | 0 | 0 | — |
case-18 | fail→pass | 14,137 | 4,476 | -68% | 1 | 1 | 0% | 2,204 | 1,888 | -14% | 0 | 0 | — |
case-19 | fail→pass | 13,902 | 5,617 | -60% | 1 | 1 | 0% | 2,223 | 2,135 | -4% | 0 | 0 | — |
case-20 | fail→pass | 16,003 | 3,731 | -77% | 1 | 1 | 0% | 2,333 | 1,728 | -26% | 0 | 0 | — |
case-21 | fail→pass | 18,769 | 7,036 | -63% | 1 | 1 | 0% | 2,829 | 2,376 | -16% | 0 | 0 | — |
case-22 | fail→pass | 15,064 | 7,969 | -47% | 1 | 1 | 0% | 2,451 | 2,393 | -2% | 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 +68 percentage points is the difference between those two pass rates over the 22 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.