Add Jumpspace To A Repo
Use this workflow for a repo that is adopting Jumpspace from scratch.
1. Install
Section titled “1. Install”npm install -D @jumpspace/cli2. Initialize
Section titled “2. Initialize”npx @jumpspace/cli init --auto--auto discovers common docs roots and writes a starter .jumpspace/config.json.
Plain init keeps the conservative default. Both paths create or refresh a
managed .gitignore block for Jumpspace runtime locks, semantic caches, and
one-shot bootstrap proposal files.
3. Add Agent Guidance
Section titled “3. Add Agent Guidance”npx @jumpspace/cli add-skill --allThis adds repo-local guidance for supported coding agents.
4. Scan And Diagnose
Section titled “4. Scan And Diagnose”npx @jumpspace/cli task scannpx @jumpspace/cli task doctornpx @jumpspace/cli task auditscan writes the generated index. doctor gives operator-friendly diagnostics.
audit validates task metadata and linked files.
5. Commit The Shared State
Section titled “5. Commit The Shared State”Usually commit:
.jumpspace/config.json.gitignoreifinitadded the Jumpspace-managed ignore block- Markdown docs with
<!-- jumpspace ... -->task blocks - repo-local agent guidance files created by
add-skill - CI files created by
init --ci github, if used
The managed ignore block leaves .jumpspace/index.json and mutation logs to
your repo policy, but ignores .jumpspace/locks/,
.jumpspace/semantic-index.json, .jumpspace/semantic-lancedb/,
jumpspace-bootstrap.json, and jumpspace-bootstrap-context.json.
Optional CI Setup
Section titled “Optional CI Setup”npx @jumpspace/cli init --ci github --dry-run --jsonnpx @jumpspace/cli init --ci githubThe generated workflow is local-only Jumpspace analysis. It scans, refreshes the semantic index, renders the PR packet, and runs audit/doctor without requiring a hosted Jumpspace service.