Install any skill in seconds. Free to start, no credit card required.
Get Started Free →get an env var, fetch a secret, missing env var, missing token/API key, load secrets from Infisical, infisical. Fetch secrets from the team's Infisical workspace into the shell environment so subsequent commands can use them.
.claude/skills/devin-axis-get-env-var/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 618% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -32% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -32% | 0% |
Fetch a secret from the team's Infisical workspace into the current shell so the next command can use it.
BLOB_READ_WRITE_TOKEN.brew install infisical/get-cli/infisical.infisical user get; if it fails, run infisical login and complete the browser flow.INFISICAL_TOKEN from a machine identity; the CLI skips login when it is present..infisical.json (workspaceId: "e9f4542a-8714-46c3-a8fd-99d8cb370aeb", empty defaultEnvironment). From the repo root, infisical defaults to the dev environment slug when --env is omitted.Run from the repo root:
bashexport NAME="$(infisical secrets get NAME --plain --silent)"
NAME with the secret name.--env <slug> for a non-default environment; this repo defaults to dev.--path /some/folder when secrets are organized in folders.Run the command through Infisical so all project secrets are available only to that process:
bashinfisical run -- <command>
--plain inside command substitution, as in export NAME="$(...)".| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | pass→pass | 15,055 | 9,872 | -34% | 1 | 1 | 0% | 1,796 | 1,347 | -25% | 0 | 0 | — |
case-05 | pass→pass | 12,286 | 8,017 | -35% | 1 | 1 | 0% | 1,305 | 791 | -39% | 0 | 0 | — |
case-10 | fail→pass | 16,355 | 9,201 | -44% | 1 | 1 | 0% | 1,808 | 1,318 | -27% | 0 | 0 | — |
case-01 | fail→pass | 13,379 | 9,745 | -27% | 1 | 1 | 0% | 1,261 | 1,296 | +3% | 0 | 0 | — |
case-02 | fail→pass | 16,399 | 19,571 | +19% | 1 | 1 | 0% | 406 | 2,914 | +618% | 0 | 0 | — |
case-03 | fail→fail | 9,196 | 8,068 | -12% | 1 | 1 | 0% | 785 | 1,027 | +31% | 0 | 0 | — |
case-04 | pass→pass | 13,458 | 12,748 | -5% | 1 | 1 | 0% | 1,681 | 1,932 | +15% | 0 | 0 | — |
case-06 | pass→pass | 14,082 | 7,362 | -48% | 1 | 1 | 0% | 1,475 | 927 | -37% | 0 | 0 | — |
case-07 | fail→pass | 13,322 | 7,633 | -43% | 1 | 1 | 0% | 1,461 | 993 | -32% | 0 | 0 | — |
case-08 | fail→pass | 11,734 | 6,993 | -40% | 1 | 1 | 0% | 1,243 | 847 | -32% | 0 | 0 | — |
case-09 | pass→pass | 12,263 | 6,596 | -46% | 1 | 1 | 0% | 1,301 | 776 | -40% | 0 | 0 | — |
case-11 | pass→pass | 8,937 | 6,620 | -26% | 1 | 1 | 0% | 727 | 728 | +0% | 0 | 0 | — |
case-12 | fail→pass | 10,484 | 8,187 | -22% | 1 | 1 | 0% | 957 | 813 | -15% | 0 | 0 | — |
case-13 | fail→pass | 16,991 | 9,015 | -47% | 1 | 1 | 0% | 1,928 | 1,145 | -41% | 0 | 0 | — |
case-14 | pass→pass | 13,034 | 7,942 | -39% | 1 | 1 | 0% | 1,349 | 964 | -29% | 0 | 0 | — |
case-16 | pass→pass | 9,671 | 6,761 | -30% | 1 | 1 | 0% | 1,602 | 791 | -51% | 0 | 0 | — |
case-17 | pass→pass | 11,244 | 8,474 | -25% | 1 | 1 | 0% | 1,130 | 1,160 | +3% | 0 | 0 | — |
case-18 | fail→pass | 14,787 | 8,737 | -41% | 1 | 1 | 0% | 1,816 | 906 | -50% | 0 | 0 | — |
case-19 | pass→pass | 14,461 | 7,923 | -45% | 1 | 1 | 0% | 1,420 | 1,030 | -27% | 0 | 0 | — |
case-20 | pass→pass | 13,032 | 7,770 | -40% | 1 | 1 | 0% | 1,350 | 1,027 | -24% | 0 | 0 | — |
case-21 | pass→pass | 13,700 | 7,586 | -45% | 1 | 1 | 0% | 1,547 | 905 | -41% | 0 | 0 | — |
case-22 | pass→pass | 7,276 | 6,623 | -9% | 1 | 1 | 0% | 345 | 705 | +104% | 0 | 0 | — |
case-23 | pass→pass | 17,331 | 10,423 | -40% | 1 | 1 | 0% | 1,950 | 1,275 | -35% | 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 22 counted toward the lift figure. The other 1 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 +35 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.