Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create a minimal working BambooHR example — fetch employee directory and single employee. Use when starting a new BambooHR integration, testing your setup, or learning basic BambooHR REST API patterns. Trigger with phrases like "bamboohr hello world", "bamboohr example", "bamboohr quick start", "simple bamboohr code", "first bamboohr call".
.claude/skills/jeremylongshore-bamboohr-hello-world/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 61% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 47% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 76% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 44% | 0% |
Verify routing, TLS, authentication, and basic permission with the smallest useful call. Do not begin with the employee directory: it contains personal data and is unnecessary for a connectivity test.
Use the tenant-local host https://company-subdomain.bamboohr.com. The official OpenAPI operation GET /api/v1/company_information returns basic company profile information and supports OAuth or API-key authentication.
Prefer a short-lived OAuth access token for partner integrations. For an internal API key, Basic auth uses the API key as username and x as password. Load secrets from the current process environment or an approved secret manager; never place a token directly in the command, URL, source file, or receipt.
ports, and dots in the subdomain input.
user-agent, request-ID, and redaction controls.
request:
bash curl --fail-with-body --silent --show-error \ --max-time 20 --output /dev/null --write-out '%{http_code}\n' \ --user "${BAMBOOHR_API_KEY}:x" \ "https://${BAMBOOHR_COMPANY_SUBDOMAIN}.bamboohr.com/api/v1/company_information"
Authorization: Bearer … from the application's secret-aware HTTP client rather than echoing a token through shell history.
server request ID when available. Do not retain the response body.
scoped test with an explicitly minimized field list.
Use Read, Glob, and Grep to find current client and configuration conventions. Use Write/Edit only to add an approved example or test. Bash is restricted to curl; show the exact host and body-retention behavior before executing it.
The optional live request can reach sensitive HR infrastructure. It is limited to HTTPS, one read-only company-information endpoint, a finite timeout, no body retention, and operator approval for the tenant and credential.
Do not make a live request, access an employee endpoint, install a package, or write credentials without explicit approval. A successful 200 proves only this identity can perform this operation; it does not prove broader access.
Return a PASS/FAIL receipt with tenant alias, endpoint class, auth mode, status, latency, request ID, redaction confirmation, and the next narrow test.
401: validate credential type and freshness without printing it.403: the identity lacks the operation; do not escalate permissions silently.429, 504, or 598: report the transient condition and defer to the retry skill.Read official evidence before running the smoke test.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,972 | 8,408 | -16% | 1 | 1 | 0% | 2,181 | 3,314 | +52% | 0 | 0 | — |
case-02 | fail→pass | 10,743 | 8,811 | -18% | 1 | 1 | 0% | 2,209 | 3,553 | +61% | 0 | 0 | — |
case-03 | fail→pass | 9,922 | 7,084 | -29% | 1 | 1 | 0% | 2,021 | 2,979 | +47% | 0 | 0 | — |
case-04 | fail→pass | 7,755 | 6,819 | -12% | 1 | 1 | 0% | 1,641 | 2,894 | +76% | 0 | 0 | — |
case-05 | pass→pass | 8,009 | 5,918 | -26% | 1 | 1 | 0% | 1,611 | 2,797 | +74% | 0 | 0 | — |
case-06 | pass→pass | 6,979 | 3,934 | -44% | 1 | 1 | 0% | 1,153 | 2,191 | +90% | 0 | 0 | — |
case-07 | pass→pass | 8,461 | 5,251 | -38% | 1 | 1 | 0% | 1,576 | 2,510 | +59% | 0 | 0 | — |
case-08 | pass→pass | 4,631 | 3,146 | -32% | 1 | 1 | 0% | 855 | 2,106 | +146% | 0 | 0 | — |
case-09 | pass→pass | 3,295 | 1,787 | -46% | 1 | 1 | 0% | 627 | 1,821 | +190% | 0 | 0 | — |
case-10 | pass→pass | 12,263 | 6,719 | -45% | 1 | 1 | 0% | 2,304 | 2,770 | +20% | 0 | 0 | — |
case-11 | pass→pass | 6,791 | 3,697 | -46% | 1 | 1 | 0% | 1,145 | 2,161 | +89% | 0 | 0 | — |
case-12 | pass→pass | 9,011 | 3,587 | -60% | 1 | 1 | 0% | 1,684 | 2,050 | +22% | 0 | 0 | — |
case-13 | pass→pass | 10,252 | 10,534 | +3% | 1 | 1 | 0% | 1,763 | 3,242 | +84% | 0 | 0 | — |
case-14 | fail→pass | 8,652 | 3,574 | -59% | 1 | 1 | 0% | 1,462 | 2,099 | +44% | 0 | 0 | — |
case-15 | pass→pass | 7,385 | 4,824 | -35% | 1 | 1 | 0% | 1,316 | 2,248 | +71% | 0 | 0 | — |
case-16 | pass→pass | 5,970 | 2,654 | -56% | 1 | 1 | 0% | 1,156 | 2,028 | +75% | 0 | 0 | — |
case-17 | pass→pass | 7,699 | 2,381 | -69% | 1 | 1 | 0% | 1,576 | 1,911 | +21% | 0 | 0 | — |
case-18 | fail→pass | 6,893 | 3,907 | -43% | 1 | 1 | 0% | 1,387 | 2,261 | +63% | 0 | 0 | — |
case-19 | fail→pass | 10,236 | 5,293 | -48% | 1 | 1 | 0% | 2,131 | 2,584 | +21% | 0 | 0 | — |
case-20 | pass→pass | 8,612 | 9,287 | +8% | 1 | 1 | 0% | 1,779 | 3,469 | +95% | 0 | 0 | — |
case-21 | pass→pass | 13,288 | 11,871 | -11% | 1 | 1 | 0% | 2,754 | 3,994 | +45% | 0 | 0 | — |
case-22 | pass→pass | 10,161 | 9,906 | -3% | 1 | 1 | 0% | 2,063 | 3,641 | +76% | 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 +32 percentage points is the difference between those two pass rates over the 22 comparable cases.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
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.