ry-initlisted

# ry-init ## Purpose Build a verified mental model of the requested project scope before implementation. If the scope is a sphere such as backend or mobile UI, inspect the entire sphere and all integration points needed to understand how it works end to end. `ry-init` is read-only for project knowledge by default. It may restore or bootstrap agent-only context from `fullrepo`, but it must not create, edit, or delete Serena memories unless the user explicitly asks to update/synchronize memories or an active stale-memory hook requires synchronization. ## Workflow 1. Run `bash ${CLAUDE_PLUGIN_ROOT}/scripts/git_sync_audit.sh` when available. 2. Inspect dirty work, old branches, and worktrees. If code is correct and consistent, synchronize it into `main`, push, and remove merged branches/worktrees. If risky, explain the issue in Russian and ask the user with concrete options. 3. Resolve effective project policy with `python3 ${CLAUDE_PLUGIN_ROOT}/scripts/project_flow_policy.py --json` when available. Bootstrap project agent-only context only when `fullrepo.mode` allows it. `--bootstrap-init` restores existing fullrepo context; it must not create a missing fullrepo branch unless policy sets `create_if_missing=true` or the user explicitly asks in the current turn. 4. Read `${CLAUDE_PLUGIN_ROOT}/references/init-context-pack.md` and use it as the required context-pack contract. 5. Use `serena-code-workflow`: check onboarding, list memories, read relevant memories, and use Serena