python-best-practices

Add consistent plain-language FICC glossary entries with a public source and the project day on which each term was introduced.

Replace an existing mock implementation with a real one without breaking its interface, contracts, callers, tests, progress tracking, or documentation.

Add a new deterministic Tool Layer capability end to end, including analytics, contract, governed API exposure, tests, and documentation.

Explain constrained portfolio optimization outputs in PM language without turning an allocation proposal into an executable trade instruction.

Combine portfolio exposure, rolling volatility, and maximum drawdown into a concise PM-style risk summary grounded only in tool outputs.

Interactive, source-grounded learning for the Agentic PM Lab.

Design canvas actions so Copilot and UI controls share validated handlers, durable state, explicit errors, and governed backend access.

Add consistent, executable evaluation cases with explicit routing, tool, argument, context, fact, safety, and answer expectations.

Run and explain deterministic rates and credit shock scenarios with explicit assumptions, provenance, and limitations.

Change local identity assignments or Cedar permissions without creating a second authority or weakening denied paths.

Scaffold a complete repository skill package with synchronized frontmatter, contract, example, and tests.

Run local schema, static-contract, mock-execution, and behavioral checks against one repository skill package.