rsc-harness CLI
rsc-harness is a CLI for configuring agent skills and project rules through a plan the user accepts before files change.
- 504 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The CLI writes project instructions, hooks and agent permissions; review automatic Git actions and updates separately.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/rsc-harnessDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The CLI writes project instructions, hooks and agent permissions; review automatic Git actions and updates separately.
This is third-party code. Review the repository files before installing.
What it does
rsc-harness is an agent-configuration CLI that begins by asking about the project and reading its files. The wizard lists skills and rules to install, defer or exclude. Users accept the plan before any files are written. One skill collection can serve several supported assistants. Once configured, agents read project instructions and retain information for later sessions. Session memory records Git and working-copy state without conversation text or file contents. The `reassess` command identifies changes needed after a stack change without silently installing components.
Who it is for. Technical leads maintaining shared project rules for teammates using different coding assistants.
Good fit when
- Teammates use different agents and their project instructions are drifting.
- In a fresh session, the agent misses decisions that should be stored beside the code.
- You want to inspect the reason for each skill before expanding the setup.
- After a stack change, you want to reassess instructions before adding components.
Not a fit when
- You only need a one-off answer with no project setup to maintain.
- Agent configuration and permission files must not change.
- You expect Codex to enforce the same guard hooks as Claude Code.
Example request
Prepare an rsc plan for this Codex project and show selected and deferred components before writing files.Limitations
Run the wizard inside the project you intend to configure. Claude Code hooks enforce guards for destructive commands; other clients receive instructions without that enforcement. A fresh setup allows ordinary commits, pushes and pull-request creation without asking, so review Git permissions. rsc updates within its current major version at session start. Disable automatic updates and pin a reviewed version when the environment must stay fixed.
How to disable. Use diagnostics to inspect installed files, then disable unwanted rules in the client settings. Use `rsc memory off` for memory and `rsc git-permissions off` for managed Git permissions.
FAQ
Does the first run change files?
No. It prints a plan first; writing requires another run with the accepted plan identifier.
Why is an accepted plan no longer valid?
If project evidence changes, rsc asks you to review a fresh plan before applying it.
How do I know setup is complete?
In `rsc doctor`, `Harness health` reports whether the installed setup works; `Onboarding readiness` reports whether setup is complete. A zero exit code confirms health only, so read the second line as well.
Can I undo a bad update?
Yes. `rsc restore latest` restores the latest snapshot saved before synchronization.
How long is the session journal kept?
The local journal is retained for 30 days and excluded from Git. It records repository state without conversation text.
Discussion
No questions or reviews yet.
Related
CLI toolsA CLI that keeps an existing agent on a long task: plan a step, run it, verify the result in the real environment and checkpoint progress
The /better-harness command reviews how an AI agent's work is set up in a project: goals, checks, delivery and memory, and returns a prioritized fix report
Squad creates coding-agent teams with roles and shared repository memory for distributing tasks and retaining decisions.
The Harness MCP server: one agent starts other Claude Code and Codex sessions on your computer, waits for them, reads their answers and closes them
A CLI that detects your project's stack and installs matching agent skills from a vetted registry in one command
A set of skills that hand a task to a separate CLI agent while you review the diff and land the commit yourself
README in short
Diagnostics distinguish setup health from onboarding completion. After project changes, the user reviews the skill set again; rsc installs new components after the plan is accepted.