Zvec: In-Process Vector Search and the Embedded Storage Pattern
A source audit of Alibaba Zvec, evaluating in-process vector retrieval, Proxima index kernels, single-file disk persistence, and edge RAG architectures.
Fodev JEO
LATEST INVESTIGATIONS
Source audits & field notes
Google ADK Python Deep Dive: Code-First Multi-Agent Architecture and Evaluation
An architectural source audit of Google’s Agent Development Kit (ADK) for Python, examining hierarchical delegation, sandboxed execution, and evaluation sets.
Pi Agent Architecture: Minimal Tool Primitives and Branching Execution Loops
A source audit of the Pi minimal agent runtime, evaluating 4-tool primitive kernels, tree-shaped session branching, hot-reloading extensions, and self-modification.
Plano Architecture: Decoupling Agent Inner Loops from Outer-Loop Control Planes
A source audit of Plano AI-native proxy architecture, evaluating inner-loop logic separation, outer-loop policy routing, filter-chain guardrails, and OTel tracing.
The Agent SDK Landscape: Comparing Orchestration Runtimes and Harness Primitives
An architectural comparison of OpenAI Agents SDK, Google ADK, Claude Agent SDK, and open runtimes across tool execution, handoffs, and observability.
AutoRAG-Research: Standardized Benchmarking and Modular RAG Evaluation
A source audit of AutoRAG-Research, evaluating standardized RAG benchmarking harnesses, modular retrieval-generation pipelines, and cross-dataset metrics.
DeepCode and nanobot: End-to-End Repository Synthesis from Research Papers
A source audit of HKUDS DeepCode (arXiv:2512.07921), analyzing Paper2Code pipeline architecture, blueprint distillation, and mobile chat orchestration.
Entire and the Agent-Native SDLC: Git Checkpoints, Provenance, and Intent Review
A source audit of Thomas Dohmke’s Entire platform, examining git-native session checkpoints, intent provenance, and the breakdown of traditional PR review.
EdgeQuake: High-Performance Graph-RAG in Rust for Production Retrieval
A technical source audit of EdgeQuake, analyzing its Rust implementation of LightRAG, multi-mode query routing, Apache AGE integration, and MCP server.
Harness Engineering: Architectural Principles for Scaling Agents to 1M Lines
A source audit of OpenAI’s harness engineering study, examining repo-native memory, mechanical invariant enforcement, and ephemeral observability.
인기 태그
Search article titles, categories, and tags. Full text is searched when needed.
Type to search articles.