shipshitdev avatar

Context Optimization

When your agent conversations start hitting token limits or your API bills spike from long contexts, this skill gives you the playbook for squeezing more out of what you have. It covers four main techniques: compaction (summarizing old conversation turns), observation masking (replacing verbose tool outputs with compact references), KV-cache optimization (structuring prompts to maximize prefix caching), and context partitioning (splitting work across sub-agents). The practical breakdown is solid, with specific guidance on when to mask versus summarize and how to measure if your optimizations actually work. Best for production systems where context management directly impacts cost and performance, not one-off scripts.

skills.sh installs
234
GitHub stars
24
last updated
Jun 3, 2026

Install

$ npx -y skills add shipshitdev/library --skill context-optimization --agent claude-code

Installs into .claude/skills of the current project.

Files

When your agent conversations start hitting token limits or your API bills spike from long contexts, this skill gives you the playbook for squeezing more out of what you have. It covers four main techniques: compaction (summarizing old conversation turns), observation masking (replacing verbose tool outputs with compact references), KV-cache optimization (structuring prompts to maximize prefix caching), and context partitioning (splitting work across sub-agents). The practical breakdown is solid, with specific guidance on when to mask versus summarize and how to measure if your optimizations actually work. Best for production systems where context management directly impacts cost and performance, not one-off scripts.

Browse the repository on GitHub

Sponsors

Details

Repository
github.com/shipshitdev/library
First seen
Jun 3, 2026

More from shipshitdev/library

All 56

Recommended

1juliusbrussee avatarcavemanjuliusbrussee/cavemanUltra-compressed communication mode. Cuts output tokens 65% (measured) by speaking like caveman while keeping full technical accuracy. Supports intensity levels618.3Kskills.sh installs
2mattpocock avatargrill-memattpocock/skillsA relentless interview to sharpen a plan or design.1.5Mskills.sh installs
3shadcn avatarimproveshadcn/improveSurvey any codebase as a senior advisor and produce prioritized, self-contained implementation plans for OTHER models/agents to execute. Strictly read-only on s40Kskills.sh installs
4obra avatarsystematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes326.7Kskills.sh installs
5forrestchang avatarkarpathy-guidelinesforrestchang/andrej-karpathy-skillsBehavioral guidelines to reduce common LLM coding mistakes through explicit assumptions, simplicity, and verifiable success criteria.17.9Kskills.sh installs
6vercel-labs avatarfind-skillsvercel-labs/skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express int4.3Mskills.sh installs