Codebase and Developer Context
Architectural articles exploring how to design, customize, and route codebase-specific context for AI coding agents.

An architectural essay on the hosting substrates and design principles of user-controlled agent memory. Analyzes the trade-offs of storing context vaults locally (flat-file Markdown/Git), on the web (REST/MCP over Cloudflare Workers), and on mobile devices (SwiftUI/iCloud), outlining seven core principles for durable, interoperable personal context.

An architectural essay arguing why conversational RAG memory plugins fail for AI coding agents. Contrasts similarity-based snippet recall with structured, git-versioned project documentation—exploring lean catalog discovery, in-place fact rewriting to prevent bloat, and codebase subindexes to eliminate blind exploratory search.

An article analyzing the extension of the LLM Wiki pattern to software repositories. Discusses compiling codebase structure into SQLite databases via tree-sitter AST parsing and exposing symbol relationship graphs to coding agents over MCP servers.

An architectural guide on structuring persistent codebase memory for coding agents. Details harness engineering, deterministic write boundaries, tiered context disclosure, and graph-based self-improving loops to eliminate decision re-litigation and quality collapse.
LLM Wiki vs RAG for Internal Codebase Memory: Which Approach Should You Use? (MindStudio Blog)mindstudio.aiA comparative guide analyzing the tradeoffs between RAG and flat-file LLM Wikis for codebase memory and agent context routing.

The official announcement of OpenWiki 0.2, detailing its integration of the Open Knowledge Format (OKF) specification.
Superwiki: An LLM Wiki and Task Tracker for Claude Code, Codex, and Copilot CLI (Mehmet Şerefoğlu on Medium)mehmetserefoglu.medium.comAn architectural article detailing Superwiki, an open-source agent skill suite transforming a repository's
docs/folder into an agent-maintained Markdown wiki and task tracker. Outlines directory structures (index.md,log.md,raw/,wiki/,tasks/,plans/), persistentAGENTS.mdsteering rules, subagent verification runs to minimize context costs, and integrations with Obsidian and an interactive HTML viewer.
An architectural article analyzing seven failure modes of memory loss in AI coding agents and proposing a layered memory graph. Synthesizes Claude Code session contexts, Graphify AST structural code maps, persistent Obsidian Markdown vaults, and Karpathy's LLM Wiki pattern to eliminate repetitive codebase rediscovery.
Your AI Agent Needs a Map: LLM Wiki vs README-Driven Documentation (Roman Dykyi on Medium)medium.comA comparative analysis examining how developers feed context to AI coding agents, contrasting project-mapping files (like README.md and CLAUDE.md) with compounding knowledge layers (LLM Wikis), and exploring token management, drift risks, and file-linking topologies.
