The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes npx claude-flow commands to perform system operations such as metrics collection, resource allocation, and performance profiling. These commands appear to be standard platform tools for managing infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill ingests historical data and workload profiles to inform its predictive scaling and optimization logic. This presents a potential attack surface where maliciously crafted input data could influence resource management decisions, although this is an inherent characteristic of its optimization purpose.
Ingestion points: analyzeWorkloadPatterns(historicalData), predictScaling(swarmId), and trainTimeSeriesModel(data).
Boundary markers: None explicitly defined for untrusted data input in the provided logic snippets.
Capability inventory: Access to system-level resource allocation through mcp.daa_resource_alloc and mcp.swarm_scale.
Sanitization: Not explicitly implemented in the provided code logic; the agent is expected to handle data validation as part of its internal implementation.

agent-resource-allocator