docs: add CLAUDE.md with codebase guide for AI assistants#9
Conversation
Covers uv workspace commands, SDK architecture (HMAC signing, namespace pattern, business-error handling), MCP preview/confirm write flow, registry extension pattern, and sales invoice field gotchas. https://claude.ai/code/session_016MYUtehiW9q7BzBsUsqXzM
AGENTS.md holds the general codebase guide readable by any AI coding agent (Codex CLI, Gemini CLI, etc.). CLAUDE.md keeps its required header and pulls in AGENTS.md via the @-import syntax so Claude Code gets the same content without duplication. https://claude.ai/code/session_016MYUtehiW9q7BzBsUsqXzM
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 5a9d74f229
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| uv sync --all-extras | ||
|
|
||
| # Run SDK tests | ||
| uv run --package merit-api pytest |
There was a problem hiding this comment.
When this command is run from the repository root in a fresh workspace, uv run --package merit-api selects the package environment but does not change the current directory (the uv run --help flag for that is --directory), so bare pytest discovers mcp/tests as well. Because the merit-api package environment does not include MCP-only dependencies like fastmcp, following the documented SDK test command can fail before exercising the SDK tests; use pytest merit_api/tests or run with --directory merit_api instead.
Useful? React with 👍 / 👎.
Covers uv workspace commands, SDK architecture (HMAC signing, namespace
pattern, business-error handling), MCP preview/confirm write flow,
registry extension pattern, and sales invoice field gotchas.
https://claude.ai/code/session_016MYUtehiW9q7BzBsUsqXzM