The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a researcher mode designed to perform deep information gathering using WebSearch and WebFetch tools, which ingests untrusted content from the internet into the agent's context.
Ingestion points: External websites, industry reports, and academic papers fetched via the researcher mode (SKILL.md).
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded prompts within the fetched content.
Capability inventory: The skill has access to file system modifications (coder mode), system design orchestration, and command-line execution.
Sanitization: There are no documented mechanisms for sanitizing or filtering instructions from external web data.
[EXTERNAL_DOWNLOADS]: The documentation provides multiple activation methods using npx claude-flow and npx claude-flow@alpha, which downloads and executes code from the NPM registry at runtime. It also references external resources hosted on github.com/ruvnet/claude-flow.
[COMMAND_EXECUTION]: The skill instructions heavily rely on shell execution via the npx CLI and local script execution (./claude-flow) to perform development tasks, which grants the agent the ability to run external binaries.

sparc-methodology