Files
launchdarkly__ai-tooling/evals/package.json
T
ari-launchdarkly 45fcc2553c feat(onboarding): port the onboardingV2 skill and add its eval suite (#168)
* Port onboardingV2 skill into skills/onboarding and add its eval suite

* chore(onboarding): drop files unused by the v2 flow

Remove first-flag/, references/1.8-summary.md, and references/1.9-editor-rules.md
(the v2 flow has no summary or editor-rules step) and repoint every reference to
the parent skill's Step 4.

* fix(onboarding): resolve mcp/fm contradiction and Cursor toggle name

Drop the edge case claiming mcp/fm needs no migration, which contradicted the
adjacent deprecation prompt, and name the Cursor toggle after the server the
config actually writes.

* fix(evals): score onboarding against the real trajectory
2026-09-03 06:27:19 -07:00

62 lines
4.9 KiB
JSON

{
"name": "@launchdarkly/skill-evals",
"private": true,
"type": "commonjs",
"scripts": {
"eval:configs-create": "promptfoo eval -c shared/defaults.yaml -c configs-create/promptfooconfig.yaml --env-file .env --no-cache -o configs-create/results.json",
"eval:configs-create:single": "promptfoo eval -c shared/defaults.yaml -c configs-create/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:configs-update": "promptfoo eval -c shared/defaults.yaml -c configs-update/promptfooconfig.yaml --env-file .env --no-cache -o configs-update/results.json",
"eval:configs-update:single": "promptfoo eval -c shared/defaults.yaml -c configs-update/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:agentcontrol-tools": "promptfoo eval -c shared/defaults.yaml -c agentcontrol-tools/promptfooconfig.yaml --env-file .env --no-cache -o agentcontrol-tools/results.json",
"eval:agentcontrol-tools:single": "promptfoo eval -c shared/defaults.yaml -c agentcontrol-tools/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:configs-variations": "promptfoo eval -c shared/defaults.yaml -c configs-variations/promptfooconfig.yaml --env-file .env --no-cache -o configs-variations/results.json",
"eval:configs-variations:single": "promptfoo eval -c shared/defaults.yaml -c configs-variations/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:flag-create": "promptfoo eval -c shared/defaults.yaml -c launchdarkly-flag-create/promptfooconfig.yaml --env-file .env --no-cache -o launchdarkly-flag-create/results.json",
"eval:flag-create:single": "promptfoo eval -c shared/defaults.yaml -c launchdarkly-flag-create/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:flag-command": "promptfoo eval -c shared/defaults.yaml -c launchdarkly-flag-command/promptfooconfig.yaml --env-file .env --no-cache -o launchdarkly-flag-command/results.json",
"eval:flag-command:single": "promptfoo eval -c shared/defaults.yaml -c launchdarkly-flag-command/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:flag-drift": "promptfoo eval -c shared/defaults.yaml -c launchdarkly-flag-drift/promptfooconfig.yaml --env-file .env --no-cache -o launchdarkly-flag-drift/results.json",
"eval:flag-drift:single": "promptfoo eval -c shared/defaults.yaml -c launchdarkly-flag-drift/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:should-flag-change": "promptfoo eval -c shared/defaults.yaml -c should-flag-change/promptfooconfig.yaml --env-file .env --no-cache -o should-flag-change/results.json",
"eval:should-flag-change:single": "promptfoo eval -c shared/defaults.yaml -c should-flag-change/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:flag-release": "promptfoo eval -c shared/defaults.yaml -c flag-release/promptfooconfig.yaml --env-file .env --no-cache -o flag-release/results.json",
"eval:flag-release:single": "promptfoo eval -c shared/defaults.yaml -c flag-release/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:flag-and-release-change": "promptfoo eval -c shared/defaults.yaml -c flag-and-release-change/promptfooconfig.yaml --env-file .env --no-cache -o flag-and-release-change/results.json",
"eval:flag-and-release-change:single": "promptfoo eval -c shared/defaults.yaml -c flag-and-release-change/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"eval:onboarding": "promptfoo eval -c shared/defaults.yaml -c onboarding/promptfooconfig.yaml --env-file .env --no-cache -o onboarding/results.json",
"eval:onboarding:single": "promptfoo eval -c shared/defaults.yaml -c onboarding/promptfooconfig.yaml --env-file .env --no-cache --filter-first-n 1",
"fixture:from-pr": "node scripts/build-pr-fixture.js",
"eval:all": "node scripts/aggregate.js --run",
"eval:aggregate": "node scripts/aggregate.js",
"eval:diff": "node scripts/diff-changed-skills.js",
"eval:badges": "node scripts/render-badges.js",
"test": "node --test tests/*.test.js",
"eval:view": "promptfoo view",
"eval:haiku": "node scripts/run-models.js --model=haiku",
"eval:sonnet": "node scripts/run-models.js --model=sonnet",
"eval:opus": "node scripts/run-models.js --model=opus",
"eval:matrix": "node scripts/run-models.js --models=haiku,sonnet,opus"
},
"devDependencies": {
"js-yaml": "^4.3.1",
"promptfoo": "^0.122.0",
"zod": "^3.22.4"
},
"dependencies": {
"@anthropic-ai/claude-agent-sdk": "^0.3.144"
},
"overrides": {
"@hono/node-server@1": "^1.19.15",
"adm-zip": "^0.6.0",
"brace-expansion": "^5.0.9",
"fast-uri": "^3.1.5",
"ip-address": "^10.3.1",
"mongoose": "^9.7.2",
"protobufjs@7": "^7.6.5",
"protobufjs@8": "^8.7.2",
"sharp": "^0.35.3",
"socket.io-parser": "^4.2.7",
"ws": "^8.21.3"
}
}