Real-world insights on Full Stack development, AI-assisted workflows, and software architecture.
Note: I’m a native Spanish-speaking Full Stack Developer writing in English to enhance my technical communication skills. Feedback on code, architecture, or grammar is always welcome!
When running local AI workloads and large language models (LLMs), raw compute (FLOPS) is often secondary to unified memory bandwidth and capacity.
Without sufficient VRAM or unified memory, running 70B+ parameter models at full precision becomes practically impossible without severe quantization, complex multi-GPU setups, or cloud dependencies.
I recently read an excellent analysis by Kunal Ganglani on how Apple’s M5 Max chip architecture impacts local AI execution:
🔗 Reading: Apple’s M5 Max Just Made the Case for Local AI Development
A must-read breakdown if you are designing local fallback environments for AI agents.
Every time an AI coding agent starts a new session, it performs costly file searches and full-file reads («grep-and-read archaeology») to understand your codebase architecture. This process wastes thousands of context tokens on structure discovery alone.
Instead of forcing the model to re-read files from scratch, pre-index your repository locally using knowledge graph tools. This allows the agent to query structural relationships (functions, classes, schemas, imports) directly via scoped queries.
Parses source code into a queryable knowledge graph using tree-sitter (supporting 30+ languages). Check out the Graphify repository on GitHub.
uv tool install graphifyy && graphify installA zero-maintenance, background code indexing tool stored in a local SQLite database with full-text search. Explore the CodeGraph repository on GitHub or find the npm package.
npm i -g @colbymchenry/codegraphRule of Thumb: Use Graphify if you need multimodal context, deep architectural relationships, or team MCP sharing. Use CodeGraph for set-it-and-forget-it local navigation that stays auto-synced.
When AI agents run terminal commands (npm test, git log, docker build, or pytest), the console outputs hundreds of lines of noise, progress bars, and redundant stack traces directly into the model’s context window.
This noise quickly consumes thousands of input tokens and triggers context rot.
RTK (Rust Token Killer) is an open-source proxy filter (Apache 2.0) that sits between your terminal and the AI agent to intercept and compress command output before it enters the context window.
git, pnpm, docker, cargo, pytest).brew install rtk) and integrates natively into tools like Claude Code, Cursor CLI, and GitHub Copilot.Bottom Line: If you only install one CLI optimization tool for AI development, make it RTK. It is the fastest way to slash token usage and prevent agent context rot during terminal execution.
This analysis is part of the broader framework covered in Context Engineering Tactics: A 10-Tactic Framework for AI Agents.
Every developer working with AI coding agents encounters the same issue: as conversation history grows, reasoning quality degrades and answers become unreliable. This decline is known as Context Rot.
Context engineering isn’t about fitting more into the context window—it’s about ruthlessly controlling what stays in it.
Context Engineering is the practice of actively managing, filtering, and structuring the information provided to Large Language Models (LLMs) during interactive agentic sessions.
Rather than treating the context window as an endless dumping ground for terminal logs, full codebase files, and long conversation turns, context engineering applies strict operational filters to ensure the model receives only actionable, high-density signal.
01. Keep the Context Window Short (The Absolute Token Rule)
Focus on absolute token counts, not percentages: A 1M token context window does not mean quality remains intact at 400k tokens.
02. Clear Early, Clear Often (Strict Session Hygiene)
Enforce a strict «One Task, One Session» workflow. Never keep «kitchen sink» sessions running indefinitely.
handoff.md file and resume from a fresh session.03. Short & Sharp Prompts (Precision Input Curation)
Point the agent to explicit file references rather than pasting raw code blocks into the chat prompt.
src/auth/session.ts) rather than entire directories.04. AGENTS.md Diet (Lean Persistent Rules)
Keep system instructions lean and ruthless. Audit your root and local AGENTS.md files constantly.
AGENTS.md files that load only when the agent works in that scope.05. MCP Tool Loadout Diet
Avoid loading monolithic Model Context Protocol (MCP) server configurations across all environments.
gh, psql, curl) which cost zero schema tokens.06. Skill Refinement (Lazy-Loaded Agent Skills)
Build highly specialized, modular skills rather than huge catch-all instruction manuals.
<500 lines per SKILL.md. Rely on local scripts for deterministic execution.07. Plan Files Strategy (External Memory & Handoffs)
Separate architectural research, execution planning, and code implementation into discrete phases. State stored in plan.md, specs.md, or todo.md survives session wipes.
research.md → plan.md → implementation. Start a fresh, clean chat session for each stage.08. Terse Output (Caveman)
Enforce concise, low-prose response formatting to minimize output token generation costs and accelerate turnarounds.
/caveman full configurations or run compression passes on local memory files.09. Mute the Shell (RTK)
Interception layer that strips noise, progress bars, and redundant stack traces from CLI tool execution.
brew install rtk and initialize with rtk init -g ..💡 Enterprise Adoption Note:
Terminal output ≠ Total bill: RTK cuts shell noise by ~80%, but system prompts, file reads, and history still make up the bulk of your context.
Debugging risk: Aggressive filtering can drop subtle error lines. Always keep tee mode enabled for full local logging.
Supply Chain & Security: A third-party binary reading command logs requires client approval and security review before corporate adoption.
💡 For an enterprise adoption breakdown and security caveats on RTK, check out my recent TIL: Enterprise Trade-offs of CLI Filters like RTK.
10. Kill Re-Reads (Graphify & CodeGraph)
Eliminate costly «grep-and-read» codebase discovery at the start of every agent session.
<100k tokens (degradation hits well before max limit).handoff.md.<500 lines per SKILL.md with deterministic script execution.research.md → plan.md → code (200-line plan replaces 150k tokens).Full Stack & AI Engineer with 15+ years of experience. Building robust software with Java, React, and modern AI workflows.