Search
@fabricorg/experiments-evals
preddy4690
LLM and code evaluators for Fabric Experiments: judge library, ModelClient contract, run aggregation.
context-fold
middlewatch
Deterministic, reversible in-session context compaction for the Pi coding agent — folds stale tool output out of the model's view, never out of the session.
...moremarkov-cli
kaspervdlaan
Markov CLI
flexx-cli
yashhhhhhhh
npx launcher for flexx: an agentic Solidity auditor that proves every finding with an executable Foundry PoC. Wraps the Python CLI (requires Python 3.11+ and Foundry).
...more@k2b/nessi
GitHub Actions
Minimal agent loop and provider adapters for nessi.
@bantam-hq/featherweight
GitHub Actions
Turn a screenplay PDF into Fountain, FDX, or structured JSON.
@0xsingularity/web3.js
GitHub Actions
Thin drop-in wrapper around @solana/web3.js preconfigured for the Singularity chain.
shadowaudit
darkmaster0345
Static API security scanner — detects undocumented and unauthenticated Express.js routes before they reach production
fluxmend
fluxmend Developers
Streaming FSM validation + layered repair for LLM outputs with embedded structured components
pi-context-compiler
zhenhuang
Experimental context-compiler mechanics with offline fixture replay and a capability-blocked Pi extension
@longzai-intelligence-auth/elysia
joydragon
No description available
@longzai-intelligence-auth/jwt
joydragon
No description available
@aopslabs/api-client
mzsonmez
Shared HTTP client for `aops-server` API (used by `aops-cli` + `aops-mcp`).
supersendtx
supersend
SuperSend TX transactional email API client
keepline
GitHub Actions
Typed WebSocket client that keeps the line up: reconnection with jittered backoff, heartbeats and half-open detection, schema validation, reconnect-surviving subscriptions, and a real observability stream. ESM-first, zero dependencies, with React 19 bindi
...more@velliajs/telegram
vellia
Telegram Bot Library By Valera
eve-openai-compaction
ewhauser
A Codex-style checkpoint compaction strategy for Eve agents.
n8n-nodes-openrouter-automaai
automaai
OpenRouter Chat Model para n8n com cache de prompt, telemetria de tokenUsage e campo user para rastreio de custo por cliente. Fork de n8n-nodes-openrouter-cache-chat-model.
...morellm-failover-router
Mohamed Shaban (محمد شعبان العتماني)
Resilient Multi-Provider LLM Fallback Router (DeepSeek, OpenAI, Ollama, Anthropic).
prompt-shield-guard
Mohamed Shaban (محمد شعبان العتماني)
LLM Prompt Injection Defender & Structured Output Enforcement Guardrail.