ReadyBase Team Guide
Scan every repo at once. Keep your team's AI-readiness aligned across projects. One shared score dashboard, one source of truth.
1. The team problem
Individual developers run ReadyBase on their own repos and get a score. But teams have a bigger picture:
- Which repos are AI-ready? You have 10 projects. Are they all safe for AI to touch, or are some risky?
- Where's the work? One repo has 10 known issues, another has 30. What's the priority?
- Who owns what? Bus factor is high in Project A, low in Project B. Where's the knowledge risk?
- Consistent standards. Do all repos follow the same guard rules and conventions your team agreed on?
Individual scores are useful. A portfolio view is essential for teams at scale.
2. Portfolio scanning
Scan all your repos at once with one command:
ReadyBase walks the directory, finds all git repos, scans each one independently, and builds a ranked view. You get:
- All repos sorted by AI-readiness score (highest first).
- Trend over time: did this repo get safer or riskier last month?
- The top 3 fixes for each repo, so the team knows where to start.
- A CSV export for your spreadsheet, or JSON for CI integration.
Portfolio first, then drill in
Start with the portfolio to see which repos need attention. Then run readybase score /path/to/repo on the weak ones to get the detailed breakdown and exact recommendations.
3. Shared context across your team
When ReadyBase onboards a repo, it generates CLAUDE.md and AGENTS.md. These files live in the repo and are version-controlled. Every developer on your team gets the same context, whether they're using ReadyBase solo or working together:
- Architecture guides. "This service owns authentication, don't touch it without talking to Alice."
- High-dependency files. "This file has 40 dependents. Run
readybase relatebefore editing." - Known gaps. "Bus factor is 1 on this package. We need to pair-program to fix it."
- Guard rules. These are checked by the dashboard and by Claude Code via MCP. Violations are logged.
The files update with each scan, reflecting the current state of the codebase. Commit them to git. Your team's onboarding is then automatic: clone the repo, AI assistant reads the files, AI already knows the landscape.
4. Team guard rules
Your team can define rules that apply to all repos. For example:
- "Don't edit
cmd/files without blast-radius approval." - "
internal/payment/changes need a code review from the payments team." - "Test coverage must stay above 70% on every commit."
When an AI agent (or a person) violates a guard rule, ReadyBase logs it. The log shows which files were edited, which rules were broken, and when. Your team lead can review the log from the dashboard:
Open localhost:8765, go to the Guard tab, and see which team members or AI agents broke which rules. This isn't about punishment, it's about learning: "We said don't edit this file, but the agent did. Why? Do we need to retrain the agent, or was the rule wrong?"
5. Team setup (10 minutes)
To get your team on ReadyBase:
- One person installs ReadyBase and runs
readybase go .on the primary repo. This generates the baseline CLAUDE.md and guard rules. - Commit the generated files to git. Every team member now has the same context when they clone.
- Scan the portfolio with
readybase portfolio /path/to/all/projectsto see the overall landscape. - Run the dashboard with
readybase serveto share the score and recommendations with the team. - Sync on priorities. The dashboard shows the weakest repos and the fixes each needs. Decide which ones to tackle first.
That's it. From here, each developer can use ReadyBase solo (run it locally on their own branches) or you can set up team enforcement with shared rules and audit logs.
What's next
- Solo guide, for individual developers getting started.
- Dashboard tour, see what the portfolio view looks like.
- Pair with Prefex, cut token costs while your AI code stays safe.