https://dotnetpower.github.io/agent-tokenomics/ Tokenomics and token economics guide for LLMs & AI agents: cut token costs in VS Code Copilot, speed up coding with cctx, and profile calls with tokprof. en Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/00-concepts/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/00-concepts/ Core concepts of LLM tokenomics and token economics: stateless re-sending, request anatomy, agent loop fan-out, pricing, prompt caching, and context rot. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/01-diagnose/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/01-diagnose/ LLM agent tokenomics Stage 1 (Diagnose): measure per-request input, cached, output, and reasoning tokens, fixed prompt overhead, and turn growth curves. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/02-isolate/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/02-isolate/ LLM agent tokenomics Stage 2 (Isolate): attribute token cost to unused tools, failure loops, bloated tool outputs, or history growth one variable at a time. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/03-prove/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/03-prove/ LLM agent tokenomics Stage 3 (Prove): reproduce token waste on a frozen task set and quantify cost and confidence intervals across repeat runs. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/04-improve/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/04-improve/ LLM agent tokenomics Stage 4 (Improve): apply one targeted fix per experiment — micro-skills, tool pruning, loop guards, output truncation, or prompt caching. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/05-result/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/05-result/ LLM agent tokenomics Stage 5 (Result): re-measure token cost, task pass rate (quality guardrail), and latency against baseline to decide keep or revert. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/06-golden-dataset-and-rubric/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/06-golden-dataset-and-rubric/ Design an Excel/CSV Golden Dataset for single-turn and multi-turn LLM agents, combine keyword and intent matching, and score quality with a rubric. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/case-tokeneconomizer/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/case-tokeneconomizer/ What microsoft/TokenEconomizer does and how it maps to the troubleshooting stages. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/ LLM agent tokenomics and token economics troubleshooting guide: Diagnose, Isolate, Prove, Improve, and verify quality with tokprof and golden datasets. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/tokprof/ https://dotnetpower.github.io/agent-tokenomics/agent-tokenomics/tokprof/ Install tokprof from PyPI and profile LLM agent token usage per session, turn, and call without changing agent code. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/01-how-usage-is-charged/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/01-how-usage-is-charged/ Copilot tokenomics & token economics billing guide: understand GitHub AI Credits (input, cached, output token rates) vs. legacy Premium Request multipliers. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/02-measure-usage/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/02-measure-usage/ Measure GitHub Copilot tokenomics and AI credit spend in VS Code using Agent Debug Logs, Cache Explorer, context window controls, and cctx usage telemetry. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/03-context-hygiene/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/03-context-hygiene/ Reduce GitHub Copilot token usage with context hygiene: new chats per task, /fork, /compact, focused #file attachments, workspace exclusions, and cctx. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/04-instructions-and-tools/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/04-instructions-and-tools/ Cut recurring input token overhead in VS Code GitHub Copilot by trimming always-on custom instructions (copilot-instructions.md, AGENTS.md) and tool schemas. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/05-models-and-planning/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/05-models-and-planning/ Optimize LLM token economics in VS Code Copilot: choose models by cost per completed task, plan before coding, and isolate exploration in subagents. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/06-prompt-and-output-style/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/06-prompt-and-output-style/ Shorter prompts, structured requests, and capped output - the levers on both sides of the request. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/07-language-and-tokenization/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/07-language-and-tokenization/ Why the same sentence costs different token counts in different languages, and what to do about it. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/08-agent-modes-and-budget/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/08-agent-modes-and-budget/ Ask vs Plan vs Agent, agent loop limits, reasoning effort, and cloud-agent setup. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/09-mcp-audit/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/09-mcp-audit/ Every enabled tool is a per-request tax. Audit servers by usage and scope them per workspace. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/10-checklist/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/10-checklist/ One-page Copilot tokenomics checklist: high-impact habits before, during, and after VS Code GitHub Copilot sessions to cut token and AI credit spend. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/cctx/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/cctx/ Install Code Context (cctx) from the VS Code Marketplace to cut Copilot lookup tokens, speed up coding tasks, and inspect live AIC/USD cost savings. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/ https://dotnetpower.github.io/agent-tokenomics/copilot-tokenomics/ Practical tokenomics and token economics for GitHub Copilot in VS Code: save tokens and AI credits, trim MCP overhead, and speed up coding with cctx. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/glossary/ https://dotnetpower.github.io/agent-tokenomics/glossary/ Definitions of tokenomics, token economics, prompt caching, context window, tool pruning, and LLM agent cost terms (English ↔ Korean). Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ https://dotnetpower.github.io/agent-tokenomics/ Tokenomics and token economics guide for LLMs & AI agents: cut token costs in VS Code Copilot, speed up coding with cctx, and profile calls with tokprof. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/papers/ https://dotnetpower.github.io/agent-tokenomics/papers/ 75 papers on saving tokens, ranked by venue tier and citations. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/resources/ https://dotnetpower.github.io/agent-tokenomics/resources/ Repositories, papers, and docs for saving tokens. Curated and checked for existence on 2026-11-03. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/00-concepts/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/00-concepts/ LLM 토큰 이코노믹스(Tokenomics · Token Economics, 토큰 경제학)의 핵심 개념인 stateless 재전송, 요청 구성, 에이전트 루프, 과금 차원, 프롬프트 캐싱, 컨텍스트 열화를 살펴봅니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/01-diagnose/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/01-diagnose/ LLM 에이전트 토큰 이코노믹스 1단계(진단): 요청별 입력·캐시·출력·추론 토큰, 첫 호출 고정 오버헤드, 턴별 컨텍스트 증가 곡선을 측정합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/02-isolate/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/02-isolate/ LLM 에이전트 토큰 이코노믹스 2단계(분리): 미사용 도구, 실패 루프, 과대한 도구 출력, 대화 기록 누적 중 비용 원인을 변수 하나씩 좁힙니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/03-prove/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/03-prove/ LLM 에이전트 토큰 이코노믹스 3단계(증명): 고정된 작업 집합에서 반복 실행을 통해 토큰 낭비 규모와 신뢰구간을 정량적으로 입증합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/04-improve/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/04-improve/ LLM 에이전트 토큰 이코노믹스 4단계(개선): 마이크로 스킬, 도구 가지치기, 루프 가드, 도구 출력 축소, 프롬프트 캐싱을 실험당 하나씩 적용합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/05-result/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/05-result/ LLM 에이전트 토큰 이코노믹스 5단계(결과): 기준선과 비교해 토큰 비용, 작업 성공률(품질 가드레일), 지연 시간을 재측정하고 유지 여부를 결정합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/06-golden-dataset-and-rubric/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/06-golden-dataset-and-rubric/ 단일 턴·멀티턴 에이전트의 토큰 최적화 전후 품질을 검증하기 위한 엑셀(CSV) 골든 데이터셋 설계, 키워드·의도 하이브리드 채점, 루브릭 평가 방법입니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/case-tokeneconomizer/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/case-tokeneconomizer/ microsoft/TokenEconomizer의 작동 방식과 이 가이드의 문제 해결 단계와의 관계를 설명합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/ LLM 에이전트의 토큰 이코노믹스(Token Economics, 토큰 경제학) 6단계 문제 해결 절차와 tokprof 프로파일링 및 골든 데이터셋 평가 가이드입니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/tokprof/ https://dotnetpower.github.io/agent-tokenomics/ko/agent-tokenomics/tokprof/ PyPI에서 tokprof를 설치하고 에이전트 코드 수정 없이 세션별, 턴별, 호출별 LLM 토큰 사용량을 분석합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/01-how-usage-is-charged/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/01-how-usage-is-charged/ VS Code GitHub Copilot 토큰 이코노믹스 과금 가이드: AI 크레딧 기반 사용량 과금(입력·캐시·출력 토큰 요율)과 프리미엄 요청 배수를 설명합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/02-measure-usage/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/02-measure-usage/ VS Code Agent Debug Logs, Cache Explorer, 컨텍스트 윈도우 컨트롤, cctx usage 대시보드로 Copilot 토큰과 AI 크레딧 소비를 측정합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/03-context-hygiene/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/03-context-hygiene/ 새 채팅, /fork, /compact 컨텍스트 압축, #file 첨부, 제외 설정, cctx 블록 클램핑 조회로 VS Code Copilot 토큰 낭비를 줄입니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/04-instructions-and-tools/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/04-instructions-and-tools/ 매 요청마다 반복 전송되는 사용자 지정 지침(copilot-instructions.md, AGENTS.md)과 도구 스키마 목록을 줄여 고정 토큰 세금을 낮춥니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/05-models-and-planning/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/05-models-and-planning/ 측정 결과를 바탕으로 모델을 고르고, 코딩 전에 계획하며, 범위가 정해진 작업을 위임합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/06-prompt-and-output-style/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/06-prompt-and-output-style/ 간결한 프롬프트, 구조화된 요청, 출력 제한으로 입력과 출력 비용을 줄이는 방법을 설명합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/07-language-and-tokenization/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/07-language-and-tokenization/ 같은 문장도 언어에 따라 토큰 수가 달라지는 이유와 대응 방법을 설명합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/08-agent-modes-and-budget/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/08-agent-modes-and-budget/ Ask, Plan, Agent 모드의 차이, 에이전트 루프 제한, 추론 강도, 클라우드 에이전트 설정을 설명합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/09-mcp-audit/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/09-mcp-audit/ 사용량을 기준으로 MCP 서버를 점검하고 작업 영역별로 필요한 도구만 사용합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/10-checklist/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/10-checklist/ VS Code GitHub Copilot 토큰과 AI 크레딧 비용을 낮추기 위해 세션 시작 전·중·후에 점검하는 한 페이지 토큰 이코노믹스 루틴입니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/cctx/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/cctx/ VS Code Marketplace에서 Code Context(cctx)를 설치해 코드 탐색 토큰을 줄이고, 에이전트 개발 속도를 가속하며, 실시간 AIC/USD 절감량을 확인합니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/ https://dotnetpower.github.io/agent-tokenomics/ko/copilot-tokenomics/ VS Code GitHub Copilot을 위한 실전 토큰 이코노믹스(Token Economics, 토큰 경제학): 토큰과 AI 크레딧 절약, 컨텍스트·MCP 최적화, cctx 개발 가속. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/glossary/ https://dotnetpower.github.io/agent-tokenomics/ko/glossary/ 토큰 이코노믹스(Token Economics, 토큰 경제학), 프롬프트 캐싱, 컨텍스트 윈도우, 도구 가지치기 등 LLM 에이전트 비용 용어 사전입니다(영어 ↔ 한국어). Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/ https://dotnetpower.github.io/agent-tokenomics/ko/ LLM과 AI 에이전트를 위한 토큰 이코노믹스(Tokenomics · Token Economics, 토큰 경제학) 실전 가이드: VS Code Copilot 토큰 절약, cctx 개발 가속, tokprof 프로파일링. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/papers/ https://dotnetpower.github.io/agent-tokenomics/ko/papers/ 토큰 절약과 관련된 논문 75편을 학회 등급과 인용 수 순으로 정리했습니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/ko/resources/ https://dotnetpower.github.io/agent-tokenomics/ko/resources/ 토큰을 절약하는 데 도움이 되는 리포지토리, 논문, 문서입니다. 2026-11-03에 존재 여부를 확인하여 엄선했습니다. Fri, 06 Nov 2026 03:00:00 GMT https://dotnetpower.github.io/agent-tokenomics/demo/tokprof-sample.html https://dotnetpower.github.io/agent-tokenomics/demo/tokprof-sample.html Interactive tokprof HTML Edition Dossier sample: per-session, per-turn, and per-call LLM input token breakdown, prompt inspector, and waste findings. Fri, 06 Nov 2026 03:00:00 GMT