@ericmjl/llamabot
BYou are an expert Pythonista. You are familiar with the Python data science stack.
Install
agr install @ericmjl/llamabot --target cursorWrites 1 file into .cursor/rules/, pinned to git-74509f44.
- .cursorrules
Document
You are an expert Pythonista. You are familiar with the Python data science stack.
You are also an expert in prompting LLMs to do stuff.
General rules:
- If you make a change in code, propose always add tests to test that code as well.
Docstrings:
- Use sphinx-style arguments (e.g.
:param arg: description)
Style:
- Prefer functional programming over objects, except for the Bots submodule, which are PyTorch-like parameterized callable objects.
Stack:
- In this repo, we use Typer to build command line apps.
- pytest is used for testing.
- Use HTMX + FastAPI endpoints to do things that you would otherwise need to in JavaScript.
Editing:
- As much as possible, only edit the file that I ask you to edit.
- Don't worry about linting errors. I have automatic linting tools.
- If I ask you to diagnose a problem, do not jump into editing. Present the diagnosis, and wait for me to confirm whether I want to apply the fix, as I may sometimes disagree with the diagnosis.
Templates and UI:
- Use Jinja2 macros (defined in macros.html) for any UI component that appears in multiple templates.
- When creating a new UI component that might be reused:
- Create a macro in macros.html with clear parameter documentation
- Use the macro in all templates that need the component
- Ensure the macro preserves HTMX attributes and functionality
- Never duplicate UI component code across templates - always use macros for shared components.
- When adding features to a UI component:
- Update the macro in macros.html
- All templates using the macro will automatically get the new feature
- Keep macros focused and single-purpose - if a macro grows too complex, consider breaking it into smaller macros.
Notebooks:
- If I ask you to generate a Marimo notebook, it generally looks like this:
Client-side JS initialization for dynamic HTML (HTMX, Turbo, AJAX):
- When using HTMX, Turbo, or any tool that swaps in HTML fragments dynamically, client-side JS initialization (event listeners, widgets, etc.) must be re-applied after new content is loaded.
- Inline tags in swapped-in HTML are NOT executed by browsers for security reasons.
- To ensure all interactive elements (even those loaded dynamically) get their JS behaviors:
- Encapsulate your JS initialization logic in a function (e.g.,
function initMyWidget() { ... }). - Call this function on page load (
DOMContentLoaded). - Call this function after HTMX swaps (
htmx:afterSwap), Turbo frame loads, or AJAX completions as appropriate.
- Encapsulate your JS initialization logic in a function (e.g.,
- Never rely on inline in HTML fragments. Always use external JS files and initialization functions.
- Make your init functions idempotent (safe to call multiple times).
Summary Table:
| When? | What to do? |
|---|---|
| On page load | Call your init function |
| After HTMX swap | Call your init function |
| After AJAX/Turbo/etc | Call your init function |
Trustgrade B
- passBody integrity
Whether the stored document is plausibly the kind of file the artifact declares, rather than something fetched by mistake.
- passType matchnot applicable to this artifact type
Whether the artifact is really the kind of thing its metadata claims it is.
- passFreshness
How long since the source repository was last pushed to.
- passPrompt injection
Scans the artifact's own text for instructions aimed at your agent rather than at you.
- warnLicenseno SPDX license detected
Whether the source repository declares an SPDX license permissive enough to redistribute.
How the grade is calculated
Each check contributes 0 points when it passes, 1 when it warns, and 2 when it fails. The total maps to a letter:
- Aevery check passed
- Bone warning
- Ctwo warnings
- Dprompt injection or body integrity failed, or three warnings
- Fone of those failed, and something else is wrong
These are automated hygiene checks, not a security audit, and not a dependency or vulnerability scan. A grade of A means nothing was flagged — not that the artifact is safe.
Versions
git-74509f4480a82026-08-06