Knowledge Share
Technical articles, tutorials, and insights
LLM-Driven Automated Testing Series (14): Production Engineering — Why Testing Turns Out to Be the Cheapest Line Item in This Whole Series
A code-generation workflow I documented earlier burned thousands of dollars in tokens during debugging. The testing scenarios verified across this series — AutoRestTest testing a 15-operation API for about a dime, pytest-triage capping itself at 10 model calls per run by default — are off by orders of magnitude. This article puts the evidence from 13 prior articles side by side to answer the three questions the planning doc left for last: why testing-scenario cost is what it is, where the human-in-the-loop gate actually belongs, and which scenarios should explicitly avoid LLMs altogether.
LLM-Driven Automated Testing Series (13): Flaky Tests Aren't a Rerun-Count Problem, They're a Classification Problem
pytest-rerunfailures (477 stars) and flaky (397 stars) handle test instability with the same one-liner: if it fails, rerun it a few times. That doesn't answer what the planning doc actually asks — was this failure a real bug, environment jitter, or a badly designed test? The only project actually attempting that root-cause classification with an LLM is a 2-star, less-than-two-months-old repo called pytest-triage — and its most instructive feature turns out to be the four safety invariants it wraps around AI judgment, guaranteeing AI is never allowed to affect whether a test actually passes or fails.
LLM-Driven Automated Testing Series (12): Beyond Pixel Diffs — Who's Actually Making LLMs 'Understand' UI Changes
BackstopJS, at 7.2k stars, handles rendering noise with exactly four numeric config knobs — misMatchThreshold, requireSameDimensions, ignoreAntialiasing, usePreciseMatching — which is still just fighting font-rendering jitter with pixel-level tolerance. The project that actually wires a VLM/LLM into the visual-regression judgment loop is a 23-star repo called vlmkit, whose own internal benchmark notes state outright that one model hallucinates red into red. Putting these two side by side answers, concretely, what semantic visual regression actually solves — and what new problems it introduces.
LLM-Driven Automated Testing Series (11): Why API Testing Is Actually the Most Conservative Battleground for LLMs
Keploy has 18.5k stars and its README says 'Expand API Coverage using AI' — but grep the entire open-source Go codebase for LLM/AI keywords and you'll find zero implementation code. Schemathesis, at 3.6k stars, uses no LLM at all — it finds real server 500 errors purely through Hypothesis-based property-based fuzzing. The projects that actually call an LLM API in their core logic turn out to be two sub-100-star projects: AutoRestTest and api-automation-agent. This article pins down why: structured API input/output is naturally suited to traditional fuzzing, so what can LLMs actually do here that traditional tools can't — and why does this domain stay cautious instead of going all-in the way UI automation has?
LLM-Driven Automated Testing Series (10): Self-Healing Locators — Where Heuristic Scoring Hits the Ceiling of Semantic Understanding
Healenium is the most established open-source project in self-healing locators, and its README claims it 'leverages machine learning' — but reading the code directly reveals the core mechanism is a weighted scoring formula over DOM tree similarity, with hardcoded constants for weights and thresholds, nothing to do with machine learning. More notably, its 2026 addition of an AI-based XPath generation endpoint is gated behind a hard-coded exception: 'you must have a paid hlm-ai service.' This article first cracks open exactly how Healenium's scoring formula works, then examines how MarketSquare/robotframework-selfhealing-agents — a genuinely LLM-based self-healing project — designs its multi-agent architecture, and finally answers the question the planning doc raised: where exactly does the tradeoff between healing success rate and false-positive rate actually bite.
Open Source Project #240: XERJ — Let Your Coding Agent Query an Index Instead of Reading Files, 40x Fewer Tokens Than grep
A local, Elasticsearch-API-compatible AI search engine: one command autoindexes code, docs, logs, and PDFs so a coding agent retrieves the exact passage instead of reading whole files. Measured 2.7x fewer output tokens than a grep-driven agent and 26x fewer than memory alone. Built in Rust, ships as a single static binary, Apache-2.0 license, 7,000+ Stars, a Trendshift trending repository.
LLM-Driven Automated Testing Series (09): Mobile Automation (V) — When a Generic Agent Framework Sinks Down into a Testing Scenario
MetaGPT is a 'multi-agent software company' framework built for requirements analysis and code generation, but it ships a built-in Android testing demo. Reading the source directly overturns one premise from the planning doc: this demo does not reuse MetaGPT's usual Product Manager/Engineer/QA role division — it has exactly one custom Role, reusing instead the lower-level Team/Environment/Role/Action scheduling scaffolding. More importantly, the README states plainly that it 'referenced ideas and code from AppAgent' — this piece breaks down exactly what this generic framework inherited from AppAgent, what reusing generic scaffolding actually bought it, and what it cost.
Open Source Project #239: Awesome DSH Plugin — A 5,000+ Entry Curated List for the DeepSeek Harness Plugin Ecosystem
A community-maintained curated list of plugins for DeepSeek Harness (dsh): 22 categories, 5,000+ entries, spanning everything from UI tweaks to security auditing. No ranking, no quality scores — just a check that each plugin installs and does what its description says. CC0 licensed, 19k+ Stars, with a built-in dsh-market installer and an unusually candid warning about third-party code risk.
One Open Source Project a Day (No. 238): Impeccable — Teaching Every AI Coding Tool to Have Taste
Impeccable is a design language tool for AI coding assistants, offering 24 commands and 59 deterministic detection rules covering the full design workflow — from product-context gathering and visual system construction to accessibility/performance audits. Supports Claude Code, Cursor, Codex, Gemini CLI, and a dozen other tools. The core engine is a standalone binary that runs detection without needing an LLM. Apache-2.0 License, 81.3k Stars.
One Open Source Project a Day (No. 237): DeepSeek Bot — Turning DeepSeek Harness Into a Team of Bots With Memory, Group Chat, and Task Handoff
DeepSeek Bot (DS Bot) is a bot plugin for DeepSeek Harness (DSH) that turns DSH into a team of long-lived bots — each bot is really an independent DSH session with persistent cross-session memory, supporting group chat collaboration, @-mention task handoff, and a Main Bot that oversees everything. Its design explicitly credits ideas (not code) from OpenAI Codex, OpenCode, Hermes Agent, and others. Apache-2.0 License, 1.9k Stars.
Open Source Project #236: PentAGI — A Fully Autonomous Multi-Agent Penetration Testing System With 20+ Pro Tools Built In, 28k+ Stars
An open-source, self-hosted multi-agent system for autonomous penetration testing: describe a target and a goal, a team of specialized AI agents plans the work, runs 20+ professional tools like nmap, Metasploit, and sqlmap inside an isolated Docker sandbox, remembers what it learned, and hands you a report. Go + React, MIT license, 28k+ Stars, supports 10+ LLM providers.
Open Source Project #235: REA — Turn Any AI Agent Into a Reverse Engineer, From Websites and Electron Apps Down to Native Binaries, 187k+ Stars
An MCP server that connects your AI agent to reverse-engineering tools: see a feature you like, no source required, and have the agent inspect native binaries, JS/Electron apps, .NET assemblies, and websites — returning evidence-backed findings with their limitations spelled out — then build a similar feature for your own project. TypeScript, MIT license, 187k+ Stars, works with Claude Code, Codex, Cursor and more.