tacuchi/multirepo-space
/SKILL.md
Scaffold and manage multi-repo workspaces for AI coding agents. Orchestrates workspace-setup.sh (infrastructure) and agent-factory CLI (agents). Use when user says "create workspace", "setup multirepo", "multi-repo workspace", "add repo to workspace", "remove repo from workspace", "workspace status", "check workspace health", "init multi-repo", "orchestrate repos", "link repositories". Also use when user has multiple repos and needs a central coordination point for AI agents (Claude Code, Codex, Gemini CLI, OpenCode, Crush, Warp Oz).
tools
Updated Apr 15, 2026
$ install --global
skillsauthnpx skillsauth add tacuchi/multirepo-space multirepo-spaceInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 15, 2026, 2:25 AM5.4s6 files scanned
SKILL.md
- name:
- multirepo-space
- description:
- |
multirepo-space
Tools
This SKILL orchestrates two tools:
workspace-setup.sh- bundled script for workspace infrastructure (dirs, symlinks, settings.json, AGENTS.md/CLAUDE.md)agent-factory- npm CLI for creating AI agents (npx @tacuchi/agent-factory)
Target determination
Determine the target based on the CLI executing this skill:
- Running in Claude Code →
--target claude - Running in Codex →
--target codex - Running in Gemini CLI →
--target gemini - Running in Crush →
--target crush - Running in Warp →
--target warp
Use <target> placeholder below. Replace with the correct value for the current CLI.
One target at a time. Never use --target all.
Quick routing
User intent -> Action:
- "Create/scaffold workspace" -> Full setup flow
- "Add a repo" -> Add flow
- "Remove/detach repo" -> Remove flow
- "Check/verify/health" -> Status flow
- "Create an agent" ->
agent-factory createdirectly - Single repo project -> DO NOT use this Skill
- Monorepo (Nx/Turborepo) -> DO NOT use this Skill
Pre-flight checklist
- Workspace dir exists? If not,
mkdirfirst. - New workspace or existing? ->
setupvsadd. - All repo paths absolute? Required - symlinks break with relative paths.
- Always run with
--dry-runfirst and show the user the output before executing.
Security
Write commands (setup, add, remove) perform irreversible filesystem changes:
symlinks, file overwrites (AGENTS.md, CLAUDE.md, settings.json), and rm -rf on skill dirs.
Before any write command:
- Run the command with
--dry-runand show the output to the user. - Ask for explicit approval: "Proceed with the above changes?"
- Only after approval, re-run without
--dry-run.
Never pass -y to workspace-setup.sh without prior user approval in the current session.
agent-factory create writes files; confirm with the user before invoking.
Full setup flow
Step 1: Detect stacks
npx @tacuchi/agent-factory detect <repo_path> --json -q
Capture JSON. Extract alias, primaryTech, stackCsv, verifyCommands.
Use stackCsv exactly as returned - pass full value to --stacks.
Step 2: Preview workspace infrastructure (dry-run)
bash "$SKILL_DIR/scripts/workspace-setup.sh" setup <workspace_path> <repo1> <repo2> ... \
--stacks "stackCsv1|stackCsv2|..." --target <target> --dry-run
Show output to user. Ask: "Proceed with these changes?" On approval:
bash "$SKILL_DIR/scripts/workspace-setup.sh" setup <workspace_path> <repo1> <repo2> ... \
--stacks "stackCsv1|stackCsv2|..." --target <target> -y
Creates: dirs, symlinks, AGENTS.md (universal), target-specific context file and settings.
Step 3: Create specialist agents
Confirm with the user before running. Per repo:
npx @tacuchi/agent-factory create \
--name repo-<alias> --role specialist --scope <repo_path> \
--output <workspace_path> --target <target> -q
Auto-appends -agent suffix: repo-<alias> -> repo-<alias>-agent.md.
Step 4: Create coordinator agent
Confirm with the user before running.
npx @tacuchi/agent-factory create \
--name coordinator --role coordinator --model opus \
--specialists "repo-<alias1>-agent,repo-<alias2>-agent,..." \
--repo-count <N> --output <workspace_path> --target <target> -q
--specialists list must use full names with -agent suffix.
Add repo flow
Step 1: Detect stack
npx @tacuchi/agent-factory detect <repo_path> --json -q
Step 2: Preview add (dry-run)
bash "$SKILL_DIR/scripts/workspace-setup.sh" add <workspace_path> <repo_path> \
--alias <alias> --stack-csv "<stackCsv>" --target <target> --dry-run
Show output to user. Ask: "Proceed with these changes?" On approval:
bash "$SKILL_DIR/scripts/workspace-setup.sh" add <workspace_path> <repo_path> \
--alias <alias> --stack-csv "<stackCsv>" --target <target> -y
Step 3: Create specialist
Confirm with the user before running.
npx @tacuchi/agent-factory create \
--name repo-<alias> --role specialist --scope <repo_path> \
--output <workspace_path> --target all -q
Step 4: Regenerate coordinator
Get updated specialist list (with -agent suffix) from existing symlinks.
Confirm with the user before running.
npx @tacuchi/agent-factory create \
--name coordinator --role coordinator --model opus \
--specialists "<updated_csv_list_with_agent_suffix>" \
--repo-count <updated_N> --output <workspace_path> --target all -q
Remove repo flow
Step 1: Preview removal (dry-run)
bash "$SKILL_DIR/scripts/workspace-setup.sh" remove <workspace_path> <alias> --target <target> --dry-run
Show output to user. Ask: "Proceed with these changes?" On approval:
bash "$SKILL_DIR/scripts/workspace-setup.sh" remove <workspace_path> <alias> --target <target> -y
Removes: symlink, agent files (.agents/, .claude/agents/, .gemini/agents/, .agents/skills/), regenerates docs.
Step 2: Regenerate coordinator
Get updated specialist list from remaining symlinks. Confirm with the user before running.
npx @tacuchi/agent-factory create \
--name coordinator --role coordinator --model opus \
--specialists "<updated_csv_list>" --repo-count <updated_N> \
--output <workspace_path> --target all -q
Status flow
bash "$SKILL_DIR/scripts/workspace-setup.sh" status <workspace_path> --target <target>
npx @tacuchi/agent-factory list <workspace_path>
Custom agent flow
To create additional agents (architecture, style, code-review, etc.). Confirm with the user before running.
npx @tacuchi/agent-factory create \
--name <agent-name> \
--role custom \
--description "<short description>" \
--instructions "<body text or path to .md file>" \
--model sonnet \
--output <workspace_path> \
--target all -q
If detect returns "Generic"
Ask the user: "Stack was not auto-detected for [repo]. What is the main language/framework?"
Then pass --stack-csv to the setup/add command so it persists in config.
Constraints
- Always run
--dry-runand show output to user before any write command. - Always confirm with the user before invoking
agent-factory create. - Never pass
-ywithout explicit user approval in the current session. - Always use absolute paths - symlinks break with relative.
- Do not modify
settings.jsonmanually. - Single repo or monorepo (Nx/Turborepo) -> do NOT use this skill.
Replace $SKILL_DIR with the absolute path to this skill's directory.
Agent hierarchy
coordinator-agent (opus) -> repo-*-agent (sonnet) per repo.
Coordinator delegates to specialists via Task. Specialists execute autonomously.
Compatibility
- All targets:
.agents/*.md+.agents/skills/*/SKILL.md+AGENTS.md - Claude:
.claude/agents/*.md+CLAUDE.md+.claude/settings.json - Gemini:
.gemini/agents/*.md+GEMINI.md - Crush:
.crush.json - Warp:
docs/warp-oz/environment-example.md
Related Skills
openclaw/taskflow
tools
VerifiedTrustedCommunity
Use when work should span one or more detached tasks but still behave like one job with a single owner context. TaskFlow is the durable flow substrate under authoring layers like Lobster, ACPX, plugins, or plain code. Keep conditional logic in the caller; use TaskFlow for flow identity, child-task linkage, waiting state, revision-checked mutations, and user-facing emergence.
357,764SKILL.mdUpdated Apr 10, 2026
openclaw/extensions/lobster
tools
VerifiedTrustedCommunity
# Lobster Lobster executes multi-step workflows with approval checkpoints. Use it when: - User wants a repeatable automation (triage, monitor, sync) - Actions need human approval before executing (send, post, delete) - Multiple tool calls should run as one deterministic operation ## When to use Lobster | User intent | Use Lobster? | | ------------------------------------------------------ | --------------------------
357,764SKILL.mdUpdated Apr 10, 2026
steipete/extensions/lobster
tools
VerifiedTrustedCommunity
# Lobster Lobster executes multi-step workflows with approval checkpoints. Use it when: - User wants a repeatable automation (triage, monitor, sync) - Actions need human approval before executing (send, post, delete) - Multiple tool calls should run as one deterministic operation ## When to use Lobster | User intent | Use Lobster? | | ------------------------------------------------------ | --------------------------
357,588SKILL.mdUpdated Apr 13, 2026
steipete/xurl
tools
VerifiedTrustedCommunity
A CLI tool for making authenticated requests to the X (Twitter) API. Use this skill when you need to post tweets, reply, quote, search, read posts, manage followers, send DMs, upload media, or interact with any X API v2 endpoint.
356,423SKILL.mdUpdated Apr 13, 2026