The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill relies on the execution of an external CLI tool via npx cc-mirror tasks for strategic task management and dependency tracking. While this is the core intended functionality of the orchestration workflow, invoking unversioned packages from a public registry via npx represents a standard supply chain dependency.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from user requests and sub-agent outputs, which are interpolated into the prompts of subsequent agents (e.g., in references/examples.md). This architecture presents an inherent surface for indirect prompt injection.
Ingestion points: User vision/requests and outputs from research/exploration agents are passed as context to implementation agents.
Boundary markers: The skill explicitly mandates a 'WORKER agent' preamble in all spawned agent prompts to establish clear behavioral boundaries and prevent recursive orchestration loops.
Capability inventory: Spawned worker agents are granted access to sensitive tools including Bash, Write, Edit, and Read across the filesystem.
Sanitization: The instructions do not specify any sanitization, filtering, or escaping for the data interpolated into agent prompts.

orchestration