name	coding-agent
description	Programming workflow for the runtime — read before edit, surgical changes, test-driven fixes, and efficient use of existing tools.
when_to_use	- User asks to write, edit, refactor, debug, or test code. - User asks to review a file or diff.
license	MIT

Coding Agent

Pipeline for any code task

Locate — Use list_dir, grep_search to find relevant files.
Read — Use read_file before any edit. Never edit blind or assume line numbers.
Plan — Think about the architecture before writing code.
Edit — Use replace_file_content or multi_replace_file_content. Change only the lines the task requires. Do not rewrite entire files unless completely necessary.
Verify — run_command to execute tests, linters, or syntax checks (python -m py_compile ...).
Report — Cite file:line and briefly summarize what changed.

Rules

Zero Hallucination Tolerance: Never invent variables, functions, or imports. If you don't know the exact name, use grep_search.
No speculative features, no extra abstractions, no defensive code for impossible paths.
Match the existing style of the file being edited (indentation, naming conventions).
When fixing a bug: first reproduce it with a command or test, then fix, then re-run to confirm.
When unsure of API signatures: search the repo before guessing.

Preferred tools

Read/Search: view_file, list_dir, grep_search.
Edit: replace_file_content, multi_replace_file_content, write_to_file.
Verify: run_command (for tests or syntax checks).
Git Operations: Use run_command to stage and commit when asked.

Plus depuis ce dépôt

même dépôt

ai-math-research

jmoraleses/Ey-code

Conduct advanced research in Artificial Intelligence and Mathematics, including the ability to generate and run automated tests or proofs.

2026-04-241

auto-test

jmoraleses/Ey-code

Automatic test generation and execution. Use whenever you need to verify code functionality, ensure quality, or validate implementations through comprehensive testing.

2026-04-241

ey-agent-core

jmoraleses/Ey-code

Core behavioral contract for the Ey-Code agent runtime. Use always — defines how the model coordinates reasoning and tool usage.

2026-04-241

karpathy-guidelines

jmoraleses/Ey-code

Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria.

2026-04-241

pentesting-chat

jmoraleses/Ey-code

Ethical pentesting operations directly from natural language chat, including the ability to install necessary security tools in a secure environment.

2026-04-241

project-implementation

jmoraleses/Ey-code

Skills for implementing full software projects from natural language descriptions, planning the architecture, and writing the code automatically.

2026-04-241

name	coding-agent
description	Programming workflow for the runtime — read before edit, surgical changes, test-driven fixes, and efficient use of existing tools.
when_to_use	- User asks to write, edit, refactor, debug, or test code. - User asks to review a file or diff.
license	MIT

Coding Agent

Pipeline for any code task

Locate — Use list_dir, grep_search to find relevant files.
Read — Use read_file before any edit. Never edit blind or assume line numbers.
Plan — Think about the architecture before writing code.
Edit — Use replace_file_content or multi_replace_file_content. Change only the lines the task requires. Do not rewrite entire files unless completely necessary.
Verify — run_command to execute tests, linters, or syntax checks (python -m py_compile ...).
Report — Cite file:line and briefly summarize what changed.

Rules

Zero Hallucination Tolerance: Never invent variables, functions, or imports. If you don't know the exact name, use grep_search.
No speculative features, no extra abstractions, no defensive code for impossible paths.
Match the existing style of the file being edited (indentation, naming conventions).
When fixing a bug: first reproduce it with a command or test, then fix, then re-run to confirm.
When unsure of API signatures: search the repo before guessing.

Preferred tools

Read/Search: view_file, list_dir, grep_search.
Edit: replace_file_content, multi_replace_file_content, write_to_file.
Verify: run_command (for tests or syntax checks).
Git Operations: Use run_command to stage and commit when asked.