6
v1.0.0Explore and compare multiple parallel solution designs ("universes") before coding, then converge on a single, well-justified approach. Use when problems are...
⭐ 0· 218·0 current·0 all-time
byopenlang@openlang-cn
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The skill's description (generate multiple solution 'universes' and converge on one) matches the SKILL.md instructions. It is instruction-only and asks for conceptual analysis and a plan — nothing in the metadata or manifest requests unrelated resources. Minor metadata inconsistency: registry name/slug is '6' while SKILL.md calls the skill 'multiverse-architect' — this is odd but not a security risk.
Instruction Scope
The SKILL.md expects the agent to build a problem snapshot, mention relevant tech stack, files, and conventions, and to infer constraints (team skills, deadlines) from context. This is coherent for an architecture advisor, but it is open-ended: the agent will rely on whatever project context it can access. The instructions do not tell the agent to read unrelated system files, environment variables, or external endpoints; however, effective use requires access to project/repo context (expected).
Install Mechanism
No install spec or code files are present. The skill is instruction-only, so nothing will be downloaded or written to disk by an installer.
Credentials
The skill declares no required environment variables, credentials, or config paths. The SKILL.md does not attempt to read secrets or request unrelated credentials.
Persistence & Privilege
Flags are default (always: false, user-invocable: true). The skill does not request permanent presence or elevated privileges. Autonomous invocation is allowed by platform defaults but there are no other red flags that amplify risk.
Assessment
This skill is an architecture/design advisor and is internally coherent. Before enabling it: (1) understand it will ask for and rely on project/repo context — only enable if you trust the agent to read your repository files; (2) it does not require installs or credentials, and nothing will be written by the skill itself, but review any implementation plans it outputs before applying changes; (3) note the metadata oddity (registry name '6' vs SKILL.md name 'multiverse-architect') and the lack of a homepage/source — if provenance matters to you, ask the publisher for more info. Do not provide secrets or external credentials to the skill.Like a lobster shell, security has layers — review code before you run it.
latestvk970p6213m1c286ge8d44nkjzh82ragw
License
MIT-0
Free to use, modify, and redistribute. No attribution required.