What does this skill do, and when should you use it?
This skill is designed for troubleshooting Multica runtimes or daemons: agent not running, task not claimed, runtime offline, workdir or session reuse issues, repository checkout failures. It provides a set of `multica` CLI commands to inspect agents, runtimes, and repository status, and to safely update/delete runtimes or check out repositories when needed. The skill emphasizes reading the chain before making changes to avoid disrupting active agent execution. It is suitable for administrators or operators of Multica to keep agent execution healthy.
Executes commands like multica agent get, multica runtime list, and multica repo checkout to diagnose issues. Provides multica runtime usage and multica runtime activity for runtime details. Performs write operations such as multica runtime update and multica runtime delete (with permission checks). Checks out repositories, optionally with a specific ref or branch.
- When an agent didn't run a task, check agent status, runtime online status, and task claiming.
- When a repo checkout fails, determine whether it's a task-level checkout or project resource issue.
- When needing to update or delete a runtime but concerned about affecting active agents, use cascade delete or permission controls.
- When wanting to inspect a runtime's usage or activity history.
How do you install this skill?
- The skill depends on Multica daemon and local environment, which may be network constrained, especially accessing overseas services from mainland China may be unstable, affecting repo checkout etc.
- Ensure understanding of the impact of runtime and repo commands before executing, especially runtime update and delete which may affect active tasks; deletion requires explicit --cascade flag.
- Skill is not available in Chinese, which may reduce accessibility for Chinese users.
- Shell / CLI
- Network access
- Local filesystem
Multica CLI
This skill is part of the Multica repo, bundled as a skill collection. Requires the Multica CLI. Installation methods are in the repo README (e.g., via Homebrew: brew install multica-ai/tap/multica). The skill file is at server/internal/service/builtin_skills/multica-runtimes-and-repos/SKILL.md; you can copy it into your skills directory.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/multica-ai/multica.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/server/internal/service/builtin_skills/multica-runtimes-and-repos" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
When an agent doesn't run, check in order: whether a task was created, whether the assignee is an agent or squad, whether the agent is archived, whether the runtime is online, whether the daemon heartbeat is recent, and whether the task was claimed. Use commands like multica agent get <agent-id> --output json, multica runtime list --output json, and multica repo checkout <repo-url>. For write operations (e.g., multica runtime update, multica runtime delete), confirm permissions and consider impact on active agents.
What are this skill's strengths and limitations?
- Provides a clear diagnostic procedure and commands for quick troubleshooting.
- Covers common runtime and repo management operations: update, delete, checkout.
- Emphasizes safety, such as cascade handling for active agents on runtime deletion.
- Open-source, freely usable and modifiable.
- Tied to the Multica platform; not applicable to other agent platforms.
- Some commands (update/delete) require specific permissions; ordinary users may lack them.
- The skill description does not include a full command reference; may need to consult source docs.
- No explicit test coverage or cross-platform validation; behavior may differ on macOS/Windows/Linux.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Multica Runtimes and Repos this page | 47 · Use with care | ★ 52k | 3d ago | NOASSERTION |
| Multica Squad Operations | 56 · Use with care | ★ 52k | 3d ago | NOASSERTION |
| Multica Agent Creation Skill | 52 · Use with care | ★ 52k | 3d ago | NOASSERTION |
| ORCH — AI Agent Orchestrator | 49 · Use with care | ★ 170 | 2mo ago | MIT |
| Multica Project & Resource Manager | 51 · Use with care | ★ 52k | 3d ago | NOASSERTION |
How did FollowSkills review this skill?
The skill has protections for delete operations: requires --cascade for cascading deletion and explicitly requires user confirmation or explicit parameter. Data flow is transparent: describes interaction with daemon and provides source map. However, the skill allows write operations like runtime update and delete, and is user-invocable without automatic confirmation; least privilege is not fully verified as skill does not clearly limit usage to minimal necessary. Deduct for permission control completeness.
Skill documentation is internally consistent, commands and source map correspond. But no tests or error handling details for key paths; cannot execute validation. So reliability limited to 7 due to lack of test evidence.
Target scenario is clear: troubleshooting runtime and repo issues. Skill declares boundaries: should not restart daemons or check out arbitrary repos. Trigger is described (user-invocable with tool restriction). However, environment adaptation: depends on Multica services, may be network constrained, no Chinese support. Score 10 for clear scenario but limited evidence for boundaries and environment.
Documentation is well-structured with quick start, core model, CLI, debugging steps, and references. But missing installation notes, versioning, changelog, and maintenance responsibility in skill. License is in repository but not in skill. So governance incomplete, score 8.
Skill describes how to diagnose runtime and repo issues with commands and order. But no evidence that outputs are directly usable or require rework; value claim unverified. Static review can't execute, so effectiveness capped at 7, but based on text, only 5 due to lack of examples or comparison.
Source map provided, but not independently verifiable. No tests or CI covering key paths of this skill. Score 4 for primary material only with thin coverage.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →