Complete guided onboarding to Freeplay — analyzes codebase, migrates prompts, and integrates logging. Use whenever the user says 'onboard to Freeplay', 'set up Freeplay', 'get started with Freeplay', 'integrate Freeplay', or wants end-to-end setup.
Analyze a codebase to map its LLM implementation for Freeplay migration. Examines prompts, model configs, tool use, sessions, and agent patterns. Use when user wants to understand their LLM usage, inventory prompts, plan a migration, audit LLM patterns, or before integrating with Freeplay — even if they just say 'analyze my code' or 'what LLM stuff do I have'.
Migrate prompts from code into Freeplay for version control, A/B testing, and centralized management. Use whenever the user wants to move prompts, manage templates, set up prompt management, or mentions 'migrate prompts', 'prompt versioning', 'prompt templates', or wants to organize their prompts.
Integrate Freeplay logging, tracing, and observability into an LLM application. Use whenever the user mentions recording LLM calls, adding observability, capturing traces, logging to Freeplay, monitoring LLM usage, or wants a dashboard — even if they don't say 'Freeplay'. Also use after prompt migration, or for observability-only setups.
Assess the health, completeness, and production readiness of a Freeplay project across the data flywheel. Use when the user asks about project status, wants to know if their project is ready for production, asks what's missing in their Freeplay setup, wants a project health check, or asks "what should I set up next?" Also use when the user is first connecting a project to Freeplay.
Analyze Freeplay test run results to surface insights, evaluation metrics, and test case details. Use when the user wants to review test run performance, compare two test runs, understand evaluation scores, or identify failing test cases. Do NOT use for executing new test runs (use run-test) or managing datasets (use dataset-management).
Create, manage, and curate Freeplay datasets (prompt datasets and agent datasets). Always confirm with the user before any write operations. Use when the user wants to create a new dataset, add test cases, update datasets, manage dataset content, import test data from CSV or JSONL, create golden sets, or build evaluation datasets. Do NOT use for running tests (use run-test) or analyzing test results (use test-run-analysis).
Execute Freeplay test runs to evaluate prompt templates and agents against datasets. Always confirm with the user before running — test runs make API calls to LLM providers and may incur costs. Use when the user wants to run a test, execute tests against a dataset, kick off evaluations, or trigger a test run. Do NOT use for analyzing existing test results (use test-run-analysis) or managing datasets (use dataset-management).