setup-baseline
/setup-baseline
Create, audit, or safely reconcile a repository-root AGENTS.md with the Baseline engineering contract and observed project facts. Use only when the user explicitly asks to initialize, set up, adopt, audit, or update Baseline project instructions; do not activate for ordinary software work or general questions about AGENTS.md.
- Author
- woliveiras ↗
- Repository
- baseline ↗
- Licence
- MIT
- Last repository change
- 19.08.2026
Assessment and security
- Files in this package
- 3
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 19.08.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Create, audit, or safely reconcile a repository-root AGENTS.md with the Baseline engineering contract and observed project facts. Use only when the user explicitly asks to initialize, set up, adopt, audit, or update Baseline project instructions; do not activate for ordinary software work or general questions about AGENTS.md.
Original author description
Create, audit, or safely reconcile a repository-root AGENTS.md with the Baseline engineering contract and observed project facts. Use only when the user explicitly asks to initialize, set up, adopt, audit, or update Baseline project instructions; do not activate for ordinary software work or general questions about AGENTS.md.
How it works
- Establish the project boundary
- Select the safe operation
- Write and validate
SKILL.md preview
--- name: setup-baseline description: Create, audit, or safely reconcile a repository-root AGENTS.md with the Baseline engineering contract and observed project facts. Use only when the user explicitly asks to initialize, set up, adopt, audit, or update Baseline project instructions; do not activate for ordinary software work or general questions about AGENTS.md. --- # Setup Baseline Read [the AGENTS contract](./references/agents-contract.md) completely before editing project instructions. ## Establish the project boundary 1. Resolve the repository root from the authorized workspace and inspect its worktree state. Work only on the root instruction file unless the user names another scope. 2. Read an existing root `AGENTS.md` and relevant project-local instructions, README, contribution guide, manifests, lockfiles, CI, architecture, security, release, and test documentation. Do not inspect personal or global agent configuration. 3. Derive project identity, boundaries, authoritative sources, non-obvious structure, and commands only from current repository evidence. Do not install dependencies, call a model or provider, or access another workspace to fill gaps without specific authority. ## Select the safe operation - When `AGENTS.md` is absent, create it at the repository root. - When it already conforms, leave it unchanged and report the evidence. - When it is incomplete, preserve stronger and project-specific instructions and apply the smallest task-owned diff that reconciles the missing contract. - When a material conflict would change behavior, scope, authority, security, or verification, stop and request the smallest human decision. Do not overwrite an existing file wholesale, weaken a local rule, add managed blocks, or copy Baseline's own repository instructions. Do not invent commands, paths, architecture, tools, or guarantees. ## Write and validate 1. Express the applicable semantic requirements from the contract with concise, specific, verifiable project language. Omit sections for which no durable project fact exists instead of leaving placeholders. Preserve the contract's activity-routing semantics without requiring the user to name a skill. Do not omit or generalize an applicable owner mapping. 2. Create the optional Claude adapter only under the contract's explicit condition. Never duplicate the `AGENTS.md` body into another client file. 3. Re-read every changed instruction as one contract. Check conflicts, local links, command provenance, sensitive values, scope, and the complete diff. 4. Run only authorized static or repository checks that meaningfully validate the instructions. Treat unavailable behavioral validation as a limitation, not passing evidence. Report whether the file was created, reconciled, or already conforming; list the repository evidence used, checks actually run, preserved unrelated changes, open decisions, and residual limitations. Do not stage, commit, push, publish, release, deploy, or mutate production without authority for that exact action.
Files in this package (3)
- SKILL.md 3 KB
- agents/openai.yaml 1 KB
- references/agents-contract.md 11 KB
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · woliveiras/baseline | 90 | MIT |
| wenhaochai/claude-plugins | 86 | MIT |
Database snapshot: 13.11.2026 · package version 39a7e3445944 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 90setup
Initialize or update a project's workflow preferences in its root AGENTS.md. Use only when the user explicitly requests setup or project instruction configuration. Never select this skill automatically for ordinary work. - 100init
This skill should be used when a user asks to initialize a repository for AI agents, create or update an AGENTS.md file, scaffold agent instructions for a codebase, or improve existing agent guidance. Triggers include phrases like 'create an AGENTS.md', 'set up agent… - 100bmad-project-context
Set up, adopt, refresh, or audit a repository''s agent instructions (the AGENTS.md block) so AI agents work well in that repo. Also records observed agent mistakes as pitfalls. Use when invoked by name - 100create-agents-md
Use when the user asks to create AGENTS.md, generate agent instructions for a repository, set up CLAUDE.md, or wants to document project conventions for coding agents. Also use when user says "create agents md", "setup agent docs", or "initialize repo for agents". - 100bmad-project-context
Set up, adopt, refresh, or audit a repository''s agent instructions (the AGENTS.md block) so AI agents work well in that repo. Also records observed agent mistakes as pitfalls. Use when invoked by name - 90dev-code-init
Create or update a repository's root AGENTS.md with concise, verified instructions for future coding agents. Use when the user invokes `/init`, asks to initialize repository instructions, create AGENTS.md, or refresh stale agent guidance. Do not use for git init, package…