The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation references a CLI alternative using npx @claude-flow/cli@latest, which downloads the utility from the npm registry at runtime.
[REMOTE_CODE_EXECUTION]: The provided CLI alternative utilizes npx to execute the @claude-flow/cli package directly from the remote registry.
[PROMPT_INJECTION]: The skill ingests and summarizes trace span data, creating a potential surface for indirect prompt injection.
Ingestion points: Trace span metadata retrieved via mcp__claude-flow__memory_search from the observability namespace in SKILL.md.
Boundary markers: No specific delimiters or boundary instructions are present to isolate the retrieved span metadata from the agent's synthesis logic.
Capability inventory: The skill has access to the Bash tool and various context synthesis tools.
Sanitization: No explicit validation or sanitization of the retrieved span content is performed before it is summarized and reported.

observe-trace