domains / families / agent-tooling / latent-briefing

latent-briefing

separate

This skill should be used when the user asks to \"share memory between agents\", \"KV cache compaction for multi-agent\", \"orchestrator worker context\", \"latent briefing\", \"reduce worker tokens\", \"cross-agent memory without summarization\", or discusses Attention Matching compaction, recursive language models with workers, or token explosion in hierarchical agents.

When to use

• share memory between agents
• KV cache compaction for multi-agent
• orchestrator worker context
• latent briefing
• reduce worker tokens
• cross-agent memory without summarization
• Attention Matching compaction
• recursive language models with workers

When not to use

• API-only stacks where internal KV tensors are inaccessible
• ordinary persistent memory
• entity tracking
• graph retrieval
• general multi-agent topology without representation-level state sharing
• prefix caching that does not transform KV state
• masking or budget policy that does not transform KV state

Skill page

Source