📄️ AI Traces
The AI Traces page helps us inspect AI execution history and investigate how agents, models, and tools behave during each run. From this page, we can review trace results, apply filters, open session conversations, and inspect the internal span structure of a trace.
📄️ Providers
The Providers page is where we manage the AI vendors available in Pyplan. From this page, we can review the existing providers, create new ones, update existing records, and delete providers that are no longer needed.
📄️ Pricing
The Pricing page manages token-based pricing tiers for each AI model. From this page, we can review configured tiers, filter them by model, create new tiers, edit existing ones, and remove obsolete ranges.
📄️ Models
The Models page defines which AI models are available in Pyplan and how they are associated with providers. From this page, we can review model records, filter them by provider, create or edit models, change their active status, and delete models that are no longer required.
📄️ Skills
The Skills page lets us define reusable guidance and policies that Pyplan's AI follows when it works on our applications. A skill is a short document — for example "how our interfaces should look" or "the rules for building a P&L" — that both the built-in AI agents and the MCP integration read before creating or modifying anything.
📄️ Memory
The Memory of an application is what its AI agents have learned about how to work in it: conventions to respect, sequences that must be followed, and mistakes already made there. Unlike Skills, which we write ourselves as policy, memories are accumulated — the agent picks them up while we work with it, and from then on it takes them into account without us repeating ourselves.