The Agent Skills Directory

[SAFE]: The skill consists solely of a markdown documentation file (SKILL.md) that describes how to use an external CLI tool (claude-flow). No suspicious network requests or file operations are directly executed by the skill itself.
[NO_CODE]: There are no executable scripts, binaries, or configuration files that run code included in the provided skill package.
[PROMPT_INJECTION]: The skill documentation describes a workflow that ingests and processes untrusted data (e.g., source code, API responses), which creates a surface for indirect prompt injection. Ingestion points: Processes codebase content in analysis pipelines and transformations of API responses in run chains (SKILL.md). Boundary markers: None identified in the provided usage examples. Capability inventory: The documented tool has capabilities for codebase analysis, code refactoring (file modification), and data transformation (SKILL.md). Sanitization: No sanitization or escaping mechanisms are mentioned in the documentation.

stream-chain