Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Set up Finta fundraising CRM account and configure integrations. Use when onboarding to Finta, connecting email/calendar, or configuring investor pipeline automation. Trigger with phrases like "install finta", "setup finta", "finta onboarding", "configure finta crm".
.claude/skills/jeremylongshore-finta-install-auth/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -49% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -42% | 0% |
Set up Finta fundraising CRM at trustfinta.com. Finta is a UI-first platform for managing fundraising pipelines, investor relationships, and deal rooms. Integration is via email/calendar sync and the Finta web app -- there is no public REST API.
Finta syncs with Gmail and Outlook to automatically track investor communications:
Default stages (customizable):
| Issue | Cause | Solution | |-------|-------|----------| | Email sync not working | OAuth expired | Reconnect in Settings > Integrations | | Calendar events missing | Wrong calendar selected | Select correct calendar | | CSV import fails | Wrong format | Use Finta template CSV | | Duplicate investors | Re-import | Merge duplicates in UI |
Proceed to finta-hello-world to set up your first fundraise pipeline.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 14,120 | 22,341 | +58% | 1 | 1 | 0% | 2,381 | 2,096 | -12% | 0 | 0 | — |
case-02 | fail→pass | 18,083 | 12,584 | -30% | 1 | 1 | 0% | 2,859 | 2,468 | -14% | 0 | 0 | — |
case-03 | fail→pass | 14,325 | 7,813 | -45% | 1 | 1 | 0% | 2,407 | 2,083 | -13% | 0 | 0 | — |
case-04 | fail→pass | 13,063 | 4,736 | -64% | 1 | 1 | 0% | 2,639 | 1,335 | -49% | 0 | 0 | — |
case-05 | fail→fail | 16,679 | 12,231 | -27% | 1 | 1 | 0% | 2,414 | 2,250 | -7% | 0 | 0 | — |
case-06 | fail→pass | 11,301 | 4,650 | -59% | 1 | 1 | 0% | 2,118 | 1,235 | -42% | 0 | 0 | — |
case-07 | pass→pass | 934,314 | 3,104 | -100% | 1 | 1 | 0% | 1,952 | 1,133 | -42% | 0 | 0 | — |
case-08 | pass→pass | 911,435 | 3,201 | -100% | 1 | 1 | 0% | 1,159 | 1,012 | -13% | 0 | 0 | — |
case-09 | pass→pass | 11,486 | 7,624 | -34% | 1 | 1 | 0% | 1,917 | 1,623 | -15% | 0 | 0 | — |
case-10 | fail→pass | 8,352 | 2,074 | -75% | 1 | 1 | 0% | 1,435 | 923 | -36% | 0 | 0 | — |
case-11 | pass→pass | 11,507 | 4,083 | -65% | 1 | 1 | 0% | 1,869 | 1,125 | -40% | 0 | 0 | — |
case-12 | fail→pass | 12,990 | 5,786 | -55% | 1 | 1 | 0% | 2,176 | 1,476 | -32% | 0 | 0 | — |
case-13 | pass→pass | 10,482 | 2,046 | -80% | 1 | 1 | 0% | 1,613 | 889 | -45% | 0 | 0 | — |
case-14 | pass→pass | 13,697 | 1,977 | -86% | 1 | 1 | 0% | 2,248 | 872 | -61% | 0 | 0 | — |
case-15 | fail→fail | 13,907 | 3,256 | -77% | 1 | 1 | 0% | 2,218 | 1,038 | -53% | 0 | 0 | — |
case-16 | fail→pass | 11,128 | 4,857 | -56% | 1 | 1 | 0% | 1,784 | 1,284 | -28% | 0 | 0 | — |
case-17 | fail→pass | 5,646 | 5,153 | -9% | 1 | 1 | 0% | 960 | 1,432 | +49% | 0 | 0 | — |
case-18 | pass→pass | 5,267 | 1,931 | -63% | 1 | 1 | 0% | 952 | 856 | -10% | 0 | 0 | — |
case-19 | fail→pass | 11,259 | 5,141 | -54% | 1 | 1 | 0% | 1,857 | 1,341 | -28% | 0 | 0 | — |
case-20 | pass→pass | 6,983 | 1,997 | -71% | 1 | 1 | 0% | 1,068 | 875 | -18% | 0 | 0 | — |
case-21 | pass→pass | 10,377 | 2,668 | -74% | 1 | 1 | 0% | 1,637 | 872 | -47% | 0 | 0 | — |
case-22 | fail→pass | 10,093 | 1,790 | -82% | 1 | 1 | 0% | 1,623 | 901 | -44% | 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 +50 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.