launchdarkly-labs

Create a new LaunchDarkly AI Config with variations, model configurations, and targeting. Use this skill to programmatically create AI Configs for agent or completion mode.

Configure LaunchDarkly AI Config targeting rules via API to control which users receive specific variations. Enable A/B tests, percentage splits, segment targeting, and multi-context rules.

Create, manage, and attach tools (functions) to LaunchDarkly AI Configs. Tools enable AI agents to interact with external systems, APIs, and databases.

Manage AI Config variations - add, update, retrieve, and delete variations from existing configs. Test different models, prompts, parameters, and tools within a single AI Config.

Access LaunchDarkly AI Configs through the REST API. Learn how to authenticate and perform operations not available through SDKs.

Build safe, scalable contexts for LaunchDarkly AI Configs with cardinality controls, multi-context patterns, and agent graph attributes.

Build and manage user contexts for LaunchDarkly AI Config targeting. Use this skill to create contexts with attributes for personalization, segmentation, and experimentation.

Create, track, retrieve, update, and delete custom business metrics in LaunchDarkly. Full lifecycle management via API and SDK.

Create a new LaunchDarkly feature flag in the specified project with the given key, name, description, and tags

List existing LaunchDarkly feature flags in a project, optionally filtered by tag or search query, to check for duplicates before creating new flags

Analyze a pull request diff or issue description and suggest feature flags that should be created, including recommended keys, names, and flag types

Migrate JavaScript/TypeScript/React/Node.js code from the Statsig SDK to LaunchDarkly. Use whenever the user mentions migrating off Statsig, replacing Statsig with LaunchDarkly, porting feature gates or dynamic configs to LaunchDarkly, converting StatsigUser to LDContext, or moving Statsig session replay / autocapture to LaunchDarkly Observability. Also use when a user pastes Statsig SDK code and asks for the LaunchDarkly equivalent. Handles feature gates and dynamic configs only — experiments are flagged and left in place. CRITICAL: agentic tools routinely import a stale LaunchDarkly Node SDK pulled from training data; this skill always resolves the current version via live npm lookup before writing any code.

Use when running, building, or troubleshooting the `statsig-to-ld` CLI — a tool that migrates flag definitions, targeting rules, and metric definitions from Statsig to LaunchDarkly. Triggers include Statsig→LaunchDarkly migration, the subcommands `analyze` / `flags import` / `targeting import` / `metrics convert`, API key setup for `STATSIG_CONSOLE_KEY` and `LD_API_KEY`, interpreting `migration-report.json` / `flag-import-report.json` / `targeting-import-report.json`, and lossy-targeting opt-ins via `--accept-data-loss`.

Run LaunchDarkly AutoFactory Phase 1 on the current change set — create a feature flag (targeting off), wire the behavior behind it, add guarded-release metrics + instrumentation, flag-on/flag-off tests, a release manifest, and a review verdict. Use when the user says "run AutoFactory", "/autofactory", or asks to flag/instrument their current changes.

Collaborates with humans on HTML documents through the Scribble local annotation tool. Use when you have just generated a long HTML artifact the user will review, when the user has annotations to address on an existing doc, or when you need targeted human feedback on a specific section mid-task.

Apply LaunchDarkly brand guidelines when creating any work artifact — decks, one-pagers, PDFs, HTML pages, or other designed content. Use this skill whenever Paul asks to create, build, or design any document, presentation, one-pager, training material, or visual deliverable. Trigger even if the request doesn't explicitly mention "brand" — if Paul is making something that will be seen by others (reps, stakeholders, leadership, external), this skill should apply. Also trigger when Paul says things like "make a deck", "build a one-pager", "create a training doc", or "design something for X."