agentleFS
Sign inStart free
ulises-jeremiasulises-jeremias/agent-toolkit/skills/core/project/SKILL.md

project

Skill19 starsChanged 2 months agoMITView on GitHub

Clone, index, and orchestrate multi-repo work via agent-toolkit project — symlinks, quick

We split this skill into the rules it gives an agent

3 rules. 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 32 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Never git clone by hand when project clone can do it — it handles symlinks and registry.

    line 78
  2. Very specific:

    Never assume CWD is the target repo.

    line 79
  3. Very specific:

    Do not delete repos/ entries directly.

    line 80

“Common” is measured: kept by at least 6 in 10 of the 32 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Project

Manage the local multi-repo registry (~/.ai-workspace/repos/ and projects/ symlinks) so swarms and agents can target any repo by name, not path.

When to use

  • User says "clone X", "add repo Y", "list projects", or references owner/repo without a local path.
  • Swarm needs --workspace /path/to/repo that isn't yet cloned.
  • Workspace context shows missing projects/ symlink for a repo referenced in a prompt.

Prerequisites

  • agent-toolkit with project subcommand (agent-toolkit project --help).
  • gh authenticated for private repos (gh auth status).
  • Workspace has repos/github.com/<owner>/<repo> layout.

Workflow

1. Clone and index

agent-toolkit project clone owner/my-repo          # clone + symlink projects/my-repo -> repos/github.com/owner/my-repo
agent-toolkit project clone owner/my-repo --branch feat/x
# Alias resolution: swarm's config.py also resolves_owner_repo from prompt URLs like https://github.com/owner/repo

# Verify
agent-toolkit project list                         # indexed projects
ls -l ~/.ai-workspace/projects
ls ~/.ai-workspace/repos/github.com/owner/my-repo

2. Use with swarm and other skills

# Swarm can target any indexed repo via --workspace
agent-toolkit swarm start --recipe pair --ui herdr --runner opencode --model-profile balanced --workspace ~/.ai-workspace/repos/github.com/owner/my-repo --attach "task"
# Inside a cloned repo, just run from its directory — swarm's find_repo_root uses git rev-parse --git-common-dir

# Assistant discovery before swarm
# (assistant skill: README -> docs/ -> AGENTS.md -> CONTRIBUTING -> PR templates -> Makefile/package.json -> devcontainer -> CI -> configs)

3. Scan and sync

agent-toolkit project scan                         # re-index repos/ -> projects/
agent-toolkit project list --json | jq
agent-toolkit workspace context --json | jq '.repos'

4. Multi-repo swarms

For cross-repo tasks, clone all targets first, then launch swarms per repo with handoffs referencing the correct --workspace or AGENT_TOOLKIT_SWARM_REPO env:

agent-toolkit project clone org/repo-a
agent-toolkit project clone org/repo-b
export AGENT_TOOLKIT_SWARM_REPO=org/repo-a  # swarm's store respects this for run location

Boundaries

  • Never git clone by hand when project clone can do it — it handles symlinks and registry.
  • Never assume CWD is the target repo; use --workspace or workdir per AGENTS.md.
  • Do not delete repos/ entries directly; use agent-toolkit project remove if available, or ask before manual rm.

Delegates to

NeedSkill
Workspace packs and memory for the new repoworkspace
Repo conventions before codingassistant
Swarm orchestration on the reposwarm
Branch/PR after swarm promotiongithub-cli-workflow

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free