bmad-project-context
Set up, adopt, refresh, or audit a repository's agent instructions
by bmad-code-orgFree
Description
BMad Project Context writes a repository's agent instructions: a small, verified block inside AGENTS.md that helps AI agents work well in that repo. You bring the rules you want followed, such as governance, security or coding standards, and the skill checks the repository and fills in the rest from what it finds. It works through five intents. Setup creates the block for a new repository, adopt brings an existing one under the block, refresh updates it, record notes agent behaviour the skill has observed, and audit checks the block that is there. It is conversational, and you approve every write. If the target resolves to more than one working tree, or to one you cannot commit in, the skill asks before writing. The target defaults to the project you are working in. The skill reads its settings from the project's _bmad/ folder, and if BMad is not set up there it offers to run setup first. Use it with a request to set up or audit a repository's agent instructions. It is part of the BMad Method project on GitHub, which has about 54,000 stars. The listing shows it as free. Its security audits show warnings from Gen Agent Trust Hub and Socket, and a pass from Snyk.…
Highlights
- Writes a verified AGENTS.md instruction block
- Setup, adopt, refresh, record and audit intents
- Brings your own governance and security rules in
- You approve every write
- Asks before writing to ambiguous targets
Ratings & Reviews
No ratings yet
Be the first to rate bmad-project-context.
People also visit
bmad-project-context FAQ
What does it write?
A small block inside AGENTS.md that tells AI agents how to work in the repository. The block is verified against the repo before it is written.
Does it write without asking?
No. It is conversational, and you approve every write.
What are the five intents?
Setup for a new repo, adopt for an existing one, refresh to update the block, record to note observed agent behaviour, and audit to check what is there.
What if the target is unclear?
If the target resolves to more than one working tree, or to one you cannot commit in, the skill asks you before it writes anything.
How do I install it?
Run npx skills add https://github.com/bmad-code-org/bmad-method --skill bmad-project-context.
Is it free and checked for security?
The listing shows it as free. Its security audits show warnings from Gen Agent Trust Hub and Socket, and a pass from Snyk.
What are the best bmad-project-context alternatives?
Popular alternatives to bmad-project-context include bmad-build, bmad-code-review and bmad-correct-course. See all bmad-project-context alternatives