Skill Composer

L2 编排层 - 智能技能编排引擎。自动编排多技能协作流程，分析技能依赖关系，优化执行策略（并行/串行决策）。

This skill appears to do what it says: local orchestration logic with no credential requests or network/external installers. Before installing or running it: 1) Review src/index.js locally (already provided) and run the included tests (npm install && npm test) in a sandboxed environment. 2) Note the SKILL.md lists shell/exec capabilities though the code does not use them — ask the maintainer to clarify whether the skill will ever run external commands or call external models. 3) The implementation maps some task types to skill ids (including 'skill-composer' itself), which could lead to self-references/cycles in some inputs — test with representative workloads. 4) Because the source origin is unknown, avoid giving it access to sensitive systems or credentials until you confirm provenance and intended runtime behavior. If you need higher assurance, ask the author for a signed release or run code analysis in an isolated environment.

Type: OpenClaw Skill Name: skill-composer-pro Version: 0.1.0 The skill bundle requests high-privilege capabilities, specifically 'Bash' and 'Exec' tools in SKILL.md, which are not utilized by the provided JavaScript logic in src/index.js. While the code itself is a benign task-orchestration library, the declaration of these tools in the agent's instruction set creates a significant attack surface for prompt injection; an attacker could craft a task description that tricks the agent into executing arbitrary shell commands. The package-lock.json also includes a large number of native bindings for 'unrs-resolver', which, while likely part of the OpenClaw ecosystem, adds to the complexity and risk profile of the bundle.