Moru Core · Rules for coding agents

Moru Core is a set of instructions and scripts your coding agent reads before it starts. It picks the steps a task needs, has the agent record each check, and keeps the task open until those records exist.

Agents check their work before they call it done.

Moru Core is a set of instructions and scripts your coding agent reads before it starts. It picks the steps a task needs, has the agent record each check, and keeps the task open until those records exist.

“Done” should mean checked.

Left alone, an agent can edit files and report the task finished without running the tests or reading the project's own rules. Moru Core puts a fixed path in front of it: read first, work through the steps the task calls for, record the results, and only then report.

Ask in your own words. The checks adapt to what you asked.

There are no extra commands to memorize after setup. Ask a question, ask for a small fix, or hand over a feature as usual. Questions get a direct answer right away, and code changes go through the checks they need.

Reads your project rules first

The agent opens the repository's own instruction file, such as AGENTS.md or CLAUDE.md, before anything else. Your project rules come before Moru Core's defaults.

Picks the right steps on its own

Whether you asked a question, a small fix, or a new feature, Moru Core automatically chooses which docs to read and which checks to run.

Separates your working folder automatically

Moru Core creates a separate branch and working folder on its own and makes changes there, so the folder you are working in stays untouched.

Runs tests and saves the result

Tests are run and saved with the command and its outcome. Saying a test passed without a saved result does not count.

Checks file size and safety

Before finishing, it checks the final diff, makes sure files did not grow too large, and runs Moru Guard.

Reports done only when every check passes

The last step verifies that nothing required was skipped or failed. If a check is missing or red, the task stays open.

Your part stays simple.

Write your project's rules in one file, ask for work in plain words, and step in only when a choice belongs to you.

Put your project's rules in AGENTS.md

Write build commands, test commands, and coding rules in AGENTS.md or CLAUDE.md in your repository. Moru Core always follows your project's file first.

Follow up in a few words

Say “add a test for that too” or “now commit it.” It keeps the context from what it just did and reuses tests that already passed.

Step in only when a decision is needed

If the answer is already in the code, the agent reads it instead of asking you. It stops and asks only when requirements conflict or a step—like pushing, deleting data, or adding a paid service—needs your approval.

Save what you learned for next time

When you find the cause of a tricky bug or tell the agent “in this project, do it this way,” it saves a note for the project and reads it again on future tasks.

Automatic branch & folder isolation

You do not have to create branches or copy folders by hand. Moru Core sets up an isolated branch and working folder automatically so your current files never get mixed up.

Blocks “done” when a test fails

An agent cannot gloss over a broken test. If a check fails, completion is held until the agent fixes the issue and the test passes.

Resumes right where it stopped

When a session ends halfway, Moru Core keeps the progress automatically so the next session picks up from the same spot instead of starting over.

Remembers causes per project

Confirmed bug causes, watch-outs, and project decisions are saved locally and handed to the agent automatically at the start of the next task.

Keeps files from growing into a mess

Automatically checks that new or edited code is split cleanly instead of dumping hundreds of lines into a single file.

Works across your coding agents

The same automatic workflow and safety rules run in Claude Code, Codex, Antigravity, and Gemini CLI.

Moru Guard

Moru Guard is the safety check Moru Core runs before an agent edits your project and again before it finishes. Low-risk setup like .gitignore rules for .env files and commit checks is handled automatically, while hard-to-undo changes pause and wait for your decision.

Secrets

Finds exposed API keys and tokens without printing their values, and moves them out of code, commits, and logs.

Env files

Keeps .env and .env.local out of Git automatically, and makes sure shared .env.example files contain no real keys.

Server-only keys

Stops database URLs, admin keys, and webhook secrets that belong on the server from leaking into browser or app code.

Hard-to-undo work

Database resets, file deletion scripts, migrations, production deploys, and credential changes wait for your explicit approval.

Cost

Before adding a paid service, database, or model call, the agent checks whether caching or a simpler server helper is enough.

Git pushes

Before pushing, it checks the remote URL, whether the repository is public, and which files are being sent.