
multica-projects-and-resources
Use when creating, inspecting, updating, or debugging Multica projects and their resources (githubrepo, localdirectory).
- Installs
- 49
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/multica-ai/multica --skill multica-projects-and-resources- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads multica-projects-and-resources automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Multica Projects and Resources
Quick start
Projects are durable context containers. Resources attached to a project can affect future agent tasks.
multica project list --output json
multica project get <project-id> --output json
multica project resource list <project-id> --output json
Project resources are mutated through project resource commands/endpoints. Issue comments do not create durable project resources.
Core model
A project groups work and carries durable resources. A resource is not just display metadata; it is context later injected into task briefs and .multica/project/resources.json.
A project's description is also durable context: when an issue (or a quick-create task) is bound to a project, the project description is injected into the agent's brief under ## Project Context and written to .multica/project/resources.json as project_description. Use it for project-wide rules/context that should apply to every task in the project.
Common resource types:
github_repo— durable GitHub repo context, withresource_ref.url, optional checkoutref, and optional prompt-onlydefault_branch_hint;local_directory— daemon-local path context, withresource_ref.local_path,daemon_id, optional label, and optionalexecution_mode(in_place, the default, orworktree).
CLI
multica project list --output json
multica project get <project-id> --output json
multica project create --title "<title>" --repo <github-url> --output json
multica project create --title "<title>" --start-date 2026-03-01 --due-date 2026-03-31 --output json
multica project update <project-id> --title "<title>" --output json
multica project update <project-id> --due-date 2026-04-15 --output json
multica project update <project-id> --start-date "" --output json # clear the start date
multica project status <project-id> in_progress --output json
multica project resource list <project-id> --output json
multica project resource add <project-id> --type github_repo --url <github-url> --output json
multica project resource add <project-id> --type github_repo --url <github-url> --ref <branch-or-sha> --output json
multica project resource add <project-id> --type local_directory --local-path <abs-path> --daemon-id <daemon-id> --output json
multica project resource add <project-id> --type local_directory --local-path <abs-path> --daemon-id <daemon-id> --execution-mode worktree --output json
multica project resource update <project-id> <resource-id> --execution-mode in_place --output json
multica project resource update <project-id> <resource-id> --url <new-github-url> --output json
multica project resource update <project-id> <resource-id> --ref <branch-or-sha> --output json
multica project resource remove <project-id> <resource-id> --output json
--execution-mode decides how tasks share a local_directory. in_place (default) runs the agent in the user's
directory, one task at a time; a second task waits in waiting_local_directory. worktree gives each task its own
git worktree of that repo, so tasks run concurrently and each delivers its work as a branch in the user's repo
instead of editing the working copy. Every task of one conversation shares that branch — agent/<agent>/<issue> for
an issue, agent/<agent>/chat-<session> for a chat — and each turn's worktree starts from the previous turn's work
rather than from HEAD; a task with no conversation behind it gets agent/<agent>/<task>. Continuation is decided by
an ownership record (refs/multica/local-state/<branch>, which holds the owning conversation, the snapshot of the
user's directory the branch already carries, and the branch tip it was recorded at), never by the branch name. A
same-named branch the user created — or one that no longer contains the recorded commit, i.e. deleted and recreated or
force-moved — is left alone and the task falls back to agent/<agent>/<issue>-<id>. A turn replays only what the user changed since that snapshot; when those
edits conflict with the branch's own work the worktree is handed to the agent mid-merge and the run delivers nothing
until the agent resolves it. worktree requires the path to be a git repository with at
least one commit; tasks fail with an explicit error otherwise. The gate is the local-worktree-v1 capability the
daemon advertises — not its version string — and it is checked twice: at save time, and again against the daemon that
claims each task, so a machine whose runtime cannot do worktrees gets its tasks cancelled rather than run in place.
Saving worktree is also refused (HTTP 422, code daemon_version_unsupported) while the daemon on that machine does
not advertise the capability — the fix is updating the Multica app there, then retrying. Pass an empty value to clear
it back to the default.
For github_repo, non-JSON --ref sets resource_ref.ref, the default checkout branch/tag/SHA for future tasks in that project. JSON --ref '<json>' remains the escape hatch for full payloads or resource types not covered by shortcuts.
--start-date / --due-date are optional calendar days (YYYY-MM-DD, like issue dates). On project update, pass an empty string (--start-date "") to clear a date; an unset flag leaves it untouched.
Referring to a project in a comment
A project has no MUL-123-style identifier, so writing its title as prose
produces dead text — there is nothing for the reader's client to autolink. Use
the mention-link form instead, with the project UUID from
multica project list --output json:
[Roadmap](mention://project/<project-id>)
Every client makes it navigable, with different presentation: web and desktop
render a chip carrying the project's icon and current title, while mobile
renders an ordinary link that opens the project on tap. Unlike @agent /
@squad, it is a pure link: util.MentionRe does not even include project,
so it enqueues nothing and notifies nobody — the same no-side-effect contract
as an issue mention.
Prefer this form over pasting the project's URL. Web and desktop do unfurl a bare in-app project URL into that same chip, but mobile does not — there a pasted URL is handed to the system browser and takes the reader out of the app.
When to add a resource
Add/update a project resource when the user asks for durable project context: "把这个 GitHub repo 绑到项目上", "以后都用这个 repo", "agent 总是拿不到这个项目的仓库", or "这个项目要在我的本地目录里跑".
Project resources are durable and affect future tasks. multica repo checkout
is task-local checkout state.
Debugging wrong context
multica project get <project-id> --output json.multica project resource list <project-id> --output json.- Check
github_repo.resource_ref.url, optionalref,default_branch_hint, andlocal_directory.resource_ref.daemon_id. - Updating resources is a durable mutation. After an update, listing the resource is the verification path.
- If resources match the expected task context, inspect runtime/repo checkout path next.
Side effects
Project create/update/delete/status and project resource add/update/remove mutate durable workspace state and affect future tasks. Ask before changing local_directory unless the user explicitly requested that exact local path.
More source-backed details: references/projects-and-resources-source-map.md.
FAQ
What does the multica-projects-and-resources skill do?
Use when creating, inspecting, updating, or debugging Multica projects and their resources (githubrepo, localdirectory).
How do I install multica-projects-and-resources?
Run npx skills add https://github.com/multica-ai/multica --skill multica-projects-and-resources in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/multica-projects-and-resources/ (personal) or .claude/skills/multica-projects-and-resources/ (project).
Which AI agents can use multica-projects-and-resources?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is multica-projects-and-resources safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains multica-projects-and-resources?
It is published by multica-ai in multica-ai/multica. It has 49 recorded installs.
More from multica-ai/multica
multica-working-on-issues
multica-ai/multica
Superseded by the multica-platform skill — load that instead. This file only records where the Multica issue contracts moved to.
multica-creating-agents
multica-ai/multica
Use when creating, inspecting, or debugging a Multica agent definition via the multica agent CLI or POST /api/agents. Not for assigning issues to agents that already exist, and not for runtime task prompts.
multica-squads
multica-ai/multica
Use when creating, inspecting, updating, assigning to, or debugging a Multica squad, including how leader routing picks who runs.
multica-skill-importing
multica-ai/multica
Use when asked to import or install a specific skill into this Multica workspace from a URL or slug. Not for choosing which skill the user needs; never treat a local installer such as npx skills add as the final install.
multica-runtimes-and-repos
multica-ai/multica
Use when a Multica runtime or daemon misbehaves: agent not running, task not claimed, runtime offline, workdir or session reuse, repository checkout.
Popular code quality & workflow skills
grill-me
mattpocock/skills
A relentless interview to sharpen a plan or design.
improve-codebase-architecture
mattpocock/skills
Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.
handoff
mattpocock/skills
Compact the current conversation into a handoff document for another agent to pick up.
domain-modeling
mattpocock/skills
Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.
codebase-design
mattpocock/skills
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
diagnosing-bugs
mattpocock/skills
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.