A local-first system that treats your authorized WeChat database as a read-only intelligence source, generating daily briefings and opportunity tracking through natural-language Codex skills.
LLMOps · Eval
underdogs breaking outNiubiGEO audits how LLMs describe your product and which competitors they recommend, saving every raw answer and citation as inspectable evidence.
It’s the public lobby for a proprietary AI workbench, collecting vision docs and ecosystem partnerships while keeping the runtime behind closed doors.
It stops long-running agents from drifting by forcing them to verify progress through independent roles before every new round.
SAGE scores whether an A2A agent should keep a task, recruit complementary teammates, or hand it off entirely—then updates its beliefs from execution evidence.
Reef is infrastructure that closes the loop between serving agent traffic, collecting feedback, and shipping updated weights or harnesses without downtime.
It wraps DeepSeek Harness in a Tauri shell so you can skip the Node.js, pnpm, and Docker chores entirely.
Open-source Claude skills that run SEO and generative-engine optimization on your actual Google data instead of scraped estimates.
Because coding agents will force-push to main or leak API keys unless something stops them mid-flight.
It exists to give retail A-share traders a lightweight, self-hosted alternative to bloated commercial terminals—no forced cloud data, no pretend AI stock picks.
Turns peer-reviewed research on AI-detectable narrative structure into a cross-platform agent skill that repairs fiction at the architectural level and tunes professional prose to its venue.
FrontierAgent is an open-source terminal runtime for long-horizon research and file-based work that ships its own evaluation harness.
agenttrail is a local, zero-dependency dashboard that draws a live, zoomable map of your AI agent’s actual work—file edits, plans, and tool calls—so you can see if it is making progress or just stuck.
A prompt-engineering skill that stops Codex and other agents from leaking rejected drafts into commits, PRs, and titles.
It tracks free API proxies for Claude Code and Codex, auto-updates their quotas and prices, and scripts the client setup so you don’t have to hunt for credits yourself.
PR Lens posts animated architecture and data-flow diagrams into your pull request comments so reviewers can see the blast radius before reading a line of code.
Memmy exists so switching between Cursor, Claude Code, and Codex doesn't mean starting your project history from scratch.
zoetrope reads Claude Code's JSONL transcripts and renders the session as a live, scrubbable flow graph of agents, subagents, and tool calls in your terminal or browser.
A plugin and skin pack that turns the DeepSeek Harness Web UI into a skinnable, pet-equipped control center.
A DSH plugin that parks a draggable, squishy whale in your corner to track DeepSeek API balance, daily usage, and per-turn costs in real time.


