Imported from donggiang/moose-codex (AGENTS.md). Install upstream with npx skills add donggiang/moose-codex. Copyright stays with the author.

AGENTS.md

These guidelines bias toward caution over speed. For trivial tasks, use judgment.

1. Think Before Coding

Don't assume. Don't hide confusion. Surface tradeoffs.

Before implementing:

• State your assumptions explicitly. If uncertain, ask.
• If multiple interpretations exist, present them - don't pick silently.
• If a simpler approach exists, say so. Push back when warranted.
• If something is unclear, stop. Name what's confusing. Ask.

2. Simplicity First

Minimum code that solves the problem. Nothing speculative.

• No features beyond what was asked.
• No abstractions for single-use code.
• No "flexibility" or "configurability" that wasn't requested.
• No error handling for impossible scenarios.
• If you write 200 lines and it could be 50, rewrite it.
• If existing functionality can accomplish a subtask, use it, don't reimplement it.
• If existing functionality is nearly capable of a subtask, prefer short extensions or bug fixes of it over long rewrites of it.
• Don't be afraid to cross into submodules for feature additions or bug fixes, especially if it aids in the 'Simplicity First' principle.

Ask yourself: "Would a senior engineer say this is overcomplicated?" If yes, simplify.

3. Surgical Changes

Touch only what you must. Clean up only your own mess.

When editing existing code:

• Don't "improve" adjacent code, comments, or formatting.
• Don't refactor things that aren't broken.
• Match existing style, even if you'd do it differently.
• If you notice unrelated dead code, mention it - don't delete it.

When your changes create orphans:

• Remove imports/variables/functions that YOUR changes made unused.
• Don't remove pre-existing dead code unless asked.

The test: Every changed line should trace directly to the user's request.

4. Code Comments

• Never delete correct preexisting code comments.
• Add comments when the code alone does not make clear what's happening.
• Explain every non-obvious numeric value, including tolerances, with a nearby comment that records its source or rationale rather than merely restating it.
• Add Doxygen comments for classes and class members whose purpose is not obvious from their names.
• Use ASCII characters.

5. Code Style

• Use make_range for integer range-based for loops (e.g. for (const auto i : make_range(n))) instead of raw index loops. Use index_range(container) when iterating over the indices of a container.
• Use libmesh_map_find for map lookups instead of .at().
• MOOSE requires C++17, so modern C++ constructs up through that standard are encouraged where they increase code readability. Along those lines, when a member of a structured binding is unused, bind it as _ instead of avoiding the structured binding solely because a member is unused.
• Put multiline method implementations outside class definitions.

6. Tool Use

• Prefer builtin tools over bash commands whenever possible in the vein of reducing permission prompting.
• Before building or performing verification, including running tests or invoking a pre-existing MOOSE executable, ask the user whether their MOOSE stack uses conda unless this has already been established in the conversation. If it does, ask which conda environment to activate and wait for the answer before running the command; do not use an existing binary or current shell state as a shortcut around this check.

7. Goal-Driven Execution

Define success criteria. Loop until verified.

Transform tasks into verifiable goals:

• "Add validation" → "Write tests for invalid inputs, then make them pass"
• "Fix the bug" → "Write a test that reproduces it, then make it pass"
• "Refactor X" → "Ensure tests pass before and after"

For multi-step tasks, state a brief plan:

1. [Step] → verify: [check]
2. [Step] → verify: [check]
3. [Step] → verify: [check]

Strong success criteria let you loop independently. Weak criteria ("make it work") require constant clarification.

These guidelines are working if: fewer unnecessary changes in diffs, fewer rewrites due to overcomplication, and clarifying questions come before implementation rather than after mistakes.