Manage artifact storage and context window efficiency through cache-aware packing.
npx skills add https://github.com/dbowm91/codegg --skill context-dbowm91
Please help me install this Agent Skill. Skill: context Source: https://github.com/dbowm91/codegg/tree/main/.opencode/skills/context Command: npx skills add https://github.com/dbowm91/codegg --skill context-dbowm91
This Skill addresses the challenge of managing large AI context windows by providing tools for artifact storage, output projection, and intelligent compaction of volatile data.
Use the context skill to analyze the current session usage and apply compaction to the volatile message tail.
High-intent search queries and answers about installing and using this skill.
Volatile-tail compaction works by automatically compacting old, low-value tool-result messages while preserving recovery handles. This allows long-term memory management to maintain stable context without exceeding token budgets.
You optimize AI context windows by applying cache-aware context packing, which sorts and prioritizes context blocks by tier and budget. This ensures efficient use of provider prompt-cache prefixes and reduces overall token costs.
Yes, context window packing supports provider-agnostic usage normalization by implementing deterministic tool-output projection. This ensures stable and cost-effective model interactions across different AI providers.
You analyze cache hit rates and token usage through effective-cost analysis diagnostics. This provides insights into provider prompt-cache efficiency to help optimize model performance and reduce costs.
A limitation of volatile-tail compaction is that it removes old tool-result messages from the active context window. While recovery handles are preserved for future reference, the full original message payloads are no longer directly accessible.
d3mingParse Claude Code JSONL logs to summarize sessions and identify pending tasks.
cablateAnalyze token costs and optimize cache usage for Claude Code agents.
bcastelinoDesign, implement, and orchestrate AI agents with CrewAI and LangGraph.
eyadsibaiSummarize conversation history and mask tool outputs to reduce token usage.
