lance-yichao-context-engineering

Designs production-grade context-engineering infrastructure for AI agents using Lance Martin (LangChain) and Yichao "Peak" Ji (Manus) methodologies — action-space layering, sandbox configurations, security guardrails, reversible compaction, structured summarization, KV-cache optimization, and pre-rot benchmarking. Use when designing a new agentic system from scratch, migrating a prototype to production, hitting performance degradation across long sessions, token costs scaling faster than value delivered, benchmarking agent reliability, selecting models per task, or simplifying a bloated harness. Trigger proactively whenever the user asks "why is my agent slower after 100 tool calls" or "how do I deploy this to production" — the Lance/Yichao surface is the production-deployment lane. For memory-crisis architecture, sovereign persistent memory, or TurboQuant compression theory, use nate-b-jones-context-engineering instead.

インストールへ移動

リポジトリ: Farrice/mes-knowledge-base
ソースの最終更新活動: 2026年7月18日 15:48
検出された SKILL.md の言語: 英語
スター: 0
フォーク: 1

インストール方法

デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。

Codex または Claude でインストールこの Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。

直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。

npx skills add https://github.com/Farrice/mes-knowledge-base --skill lance-yichao-context-engineering

コマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。

ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。

ソースファイルを確認

インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。

ファイルエクスプローラー

62 ファイル

SKILL.md

ソースの指示 · 読み取り専用プレビュー

検索に戻るインストールに戻る

generate

The /generate front door — in-house, pay-as-you-go creative generation (Higgsfield replacement). One command for image, video, and audio: routes each request to the right model recipe (skills/generate/models/*.json), defers to binding creative_router lanes (people → fal-people, style-family → fantastic-posters; Higgsfield fully retired 2026-08-06), quotes paid video before running, honors prompt-level budgets ("total budget $3"), runs comparison batches across models, logs every prompt beside its asset, and auto-refreshes the Asset Command Center (/assets-board). Use when the user says "generate", "make an image/video/voiceover", "create ads/assets/creative", names a model (recraft, kling, seedance, nano banana, gpt image), or asks for a multi-model comparison. Source: RoboNuggets 9C4TRbucmhQ, rebuilt on our fal-first stack.

2026-08-06

video-studio

The Edit Bay — in-house conversational video studio (VOX-style explainers, documentaries, shorts, LinkedIn video). Transcript-driven cutting, B-roll ladder, motion graphics as code, captions, music, publish-ready packages — all agent-operated, free-first. Use when Farrice says "edit this video/footage", "make an explainer/documentary/short", "cut this recording", "add B-roll/captions/music", "turn this essay/post into a video", or any video production ask. Front door for three modes: talking-head, VO-only, zero-camera (his recorded VO + assembled visuals — NO voice clone, NO TTS narration, his ruling 2026-08-06). Source: Brad Bonanno agentic edit pipeline (extractions/brad-bonanno-edit-bay/), rebuilt on the house stack: WhisperX local + edit_bay.py ffmpeg assembly + HyperFrames/Remotion graphics + /generate (fal) — NO Higgsfield (retired 2026-08-06).

2026-08-06

brad-bonanno-explainer-video-architecture

Architects technical explainer videos using Brad Bonanno's structural patterns — survives short attention spans, communicates complete systems (not just teasers), and functions as a channel-building artifact (one video pulls subscribers, not just views). Use when planning a long-form technical explainer (5-20 min) for YouTube, structuring an explanation of a complex system the user knows but can't teach yet, balancing depth vs accessibility for a technical audience, or building an evergreen video meant to compound subscriber growth over years rather than spike then die. Trigger proactively whenever the user says "explainer video", "technical explainer", "how do I explain X", "long-form YouTube", "evergreen video", or has a complex system that needs to land with a non-expert. For Instagram-format shareworthy content use brock-johnson-shareworthy-content; for short-form video patterns use a TikTok/Reels-specific skill.

2026-08-06

このリポジトリのすべてを見る

Lance Martin & Yichao "Peak" Ji - Context Engineering

Production-grade context engineering methodology for building AI agents that maintain performance across hundreds of tool calls without degradation.

Available Workflows

Workflow

Produces

Use When

production

Production Agent Infrastructure Blueprint

A comprehensive architectural specification including action-space layers, sandbox configurations, and security guardrails.

You are designing a new agentic system from scratch or migrating a prototype to a production-grade environment.

context

Context Management & Efficiency Engine

A technical implementation plan for reversible compaction, structured summarization, and KV-cache optimization.

The agent is suffering from performance degradation due to long-running sessions or high token costs.

performance

Performance Discovery & Optimization Protocol

An evaluation report detailing the pre-rot threshold, model routing logic, and a simplified architecture roadmap.

You need to benchmark agent reliability, select the right models for specific tasks, or simplify a bloated system.

Quick Reference

Genius Context: genius.md — load before any workflow

Legacy Prompts: references/_legacy-prompts/ — archived atomic prompts

Execution Prompts (structure-pure v2)

17 deterministic practitioner prompts — each carries an Output Contract, Output Skeleton, and Quality Gate. When a deliverable matches one, Read it and honor its contract instead of improvising the output shape.

LANCE MARTIN & PEAK JI — CONTEXT ARCHITECTURE DESIGN — skills/lance-yichao-context-engineering/references/prompts-v2/01-context-architecture-design.md

LANCE MARTIN & PEAK JI — COMPACTION SYSTEM GENERATOR — skills/lance-yichao-context-engineering/references/prompts-v2/02-compaction-system-generator.md

LANCE MARTIN & PEAK JI — THREE-LAYER ACTION SPACE DESIGN — skills/lance-yichao-context-engineering/references/prompts-v2/03-action-space-design.md

LANCE MARTIN & PEAK JI — AGENT-AS-TOOL ORCHESTRATOR — skills/lance-yichao-context-engineering/references/prompts-v2/04-agent-as-tool-orchestrator.md

LANCE MARTIN & PEAK JI — STRUCTURED SUMMARIZATION SCHEMA DESIGNER — skills/lance-yichao-context-engineering/references/prompts-v2/05-summarization-schema-designer.md

Future-Proof Assessment — [architecture name] — skills/lance-yichao-context-engineering/references/prompts-v2/06-future-proof-validator.md

LANCE MARTIN & PEAK JI — CONTEXT MANAGEMENT PIPELINE GENERATOR — skills/lance-yichao-context-engineering/references/prompts-v2/07-context-management-pipeline.md

LANCE MARTIN & PEAK JI — MCP INTEGRATION ARCHITECTURE — skills/lance-yichao-context-engineering/references/prompts-v2/08-mcp-integration-architecture.md

LANCE MARTIN & PEAK JI — AGENT EVALUATION SYSTEM DESIGNER — skills/lance-yichao-context-engineering/references/prompts-v2/09-evaluation-system-designer.md

LANCE MARTIN & PEAK JI — MODEL ROUTING STRATEGY DESIGNER — skills/lance-yichao-context-engineering/references/prompts-v2/10-model-routing-strategy.md

LANCE MARTIN & PEAK JI — COLLECTIVE FEEDBACK MINING SYSTEM — skills/lance-yichao-context-engineering/references/prompts-v2/11-collective-feedback-mining.md

LANCE MARTIN & PEAK JI — AGENTIC MAP-REDUCE ORCHESTRATOR — skills/lance-yichao-context-engineering/references/prompts-v2/12-agentic-map-reduce.md

LANCE MARTIN & PEAK JI — PRE-ROT THRESHOLD DISCOVERY PROTOCOL — skills/lance-yichao-context-engineering/references/prompts-v2/13-prerot-threshold-discovery.md

LANCE MARTIN & PEAK JI — AGENT SECURITY & GUARDRAIL ARCHITECTURE — skills/lance-yichao-context-engineering/references/prompts-v2/14-agent-security-guardrails.md