GitHub - ace-agent/ace: Evolve your language agent with Agentic Context Engineering (ACE) Evolve your language agent with Agentic Context Engineering (ACE) - ace-agent/ace

ACE is a prompt optimizer algorithm in Python. When you know the exact answer based on input data apply this algorithm to improve your prompt.
Similar to #GEPA or MIPRO in DSPy.
#PromptEngineering #ContextEngineering #AI #LLM
#OpenSource Apache 2.0 lic
github.com/ace-agent/ace

GitHub - tirth8205/code-review-graph: Local knowledge graph for Claude Code. Builds a persistent map of your codebase so Claude reads only what matters — 6.8× fewer tokens on reviews and up to 49× on ... Local knowledge graph for Claude Code. Builds a persistent map of your codebase so Claude reads only what matters — 6.8× fewer tokens on reviews and up to 49× on daily coding tasks. - tirth8205/cod...

Helper for #Agents for token efficiency. #MCP or #CLI:
Builds #Code #Graph: #AST & Tree-sitter. Continuous updates with changes. 19 languages including #Python, #TypeScript #JavaScript #Go #Rust #Java. Semantic & Hybrid search.
#OpenSource #RAG #Development #Programming

MIT lic

Agent Memory Benchmark — AMB An open, reproducible leaderboard for evaluating AI agent memory and retrieval systems on real-world long-context tasks.

A new Agent Memory #Benchmark #Leaderboard

Hindsight by @vectorize.bsky.social beats every framework on verified benchmarks: #LifeBench, #BEAM, #LoCoMo, #LongMemEval, #PersonaMem

#Hindsight is MIT lic, available as #CLI only or as a Container.
#OpenSource #MemoryBenchmarks #AI #Agents #RAG

x402: micropayments for the web 🌐
AI scraping strains infra ⚠️ Servers & net bandwidth cost 💸
Tiny per-request payments could fund sites & APIs

$1 ≈ 100–1000 requests
Privacy still matters 🔐

#AI #Agents #Web Repost 🔁
👉 Discuss with me in Discord:
linktr.ee/qdrddr

GitHub - VectifyAI/PageIndex: 📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG 📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG - VectifyAI/PageIndex

⭐️🔗 GitHub github.com/VectifyAI/Pa...
👉 Discuss with me in Discord: linktr.ee/qdrddr

PageIndex: not a #VectorDB replacement 📑
Best for hierarchical Table of Content indexing at ingestion

#RAG pattern:
1️⃣ VectorDB → find chunks
2️⃣ Load page
3️⃣ PageIndex → remove unrelated sections

🐍 Python
📜 MIT #OpenSource

#AI #LLM Repost 🔁
🔗 Link in first 💬⤵️

GitHub - softwaremill/elasticmq: In-memory message queue with an Amazon SQS-compatible interface. Runs stand-alone or embedded. In-memory message queue with an Amazon SQS-compatible interface. Runs stand-alone or embedded. - softwaremill/elasticmq

Amazon SQS-compatible interface: in-memory message queue system
#OpenSource Apache 2.0 lic
github.com/softwaremill...

GitHub - chroma-core/context-1-data-gen Contribute to chroma-core/context-1-data-gen development by creating an account on GitHub.

#ChromaDB Context-1: Fast 20B model.
Small Specialized Language Model for fast #agentic search trained to retrieve supporting documents for complex, multi-hop queries. X10 faster than #LLM
#VectorDB #RAG #SLM #Agents
#OpenSource Apache 2.0 lic
github.com/chroma-core/...

Netbird: Open Source Alternative to Twingate, Cisco Secure Access and Zerotier Open source platform combining WireGuard overlay networks with Zero Trust access controls. Features SSO, MFA, device posture checks, and granular policies for secure remote connectivity.

★ Just published: Netbird — Zero Trust networking with WireGuard-based secure connectivity

Open source platform combining WireGuard overlay networks with Zero Trust access controls. Features SSO, MFA, device posture checks, and granular policies for secure remote connectivity.

GitHub - Dokploy/dokploy: Open Source Alternative to Vercel, Netlify and Heroku. Open Source Alternative to Vercel, Netlify and Heroku. - Dokploy/dokploy

#OpenSource alternative to #Vercel

github.com/Dokploy/dokp...

When you asking to calculate number of “r”s in “strawberry” LLM struggles. But there’s a trick: if you ask LLM to write & execute JS code instead that calculates the letters then you get deterministic & reliable answer. BOA opens up such possibilities providing a sandbox to test and run JS code.

(2/2) For example, LLMs not particularly good at math, but can write JS code when executed would produce perfectly correct results. Or doing Code Mode that progressively searches MCP Server tools with produced code instead of LLM tool calls, saving tokens and improving accuracy at the same time.

(1/2) This allows the LLM/Agent to generate code and test it by running it. It paves the way for Neurosymbolic AI, where an LLM generates code that requires deterministic behavior and executes it to complete the task, rather than doing so with the LLM.

GitHub - boa-dev/boa: Boa is an embeddable Javascript engine written in Rust. Boa is an embeddable Javascript engine written in Rust. - boa-dev/boa

@boajs.dev #JavaScript engine - Parses, interprets, and executes #JS #Code entirely in‑process. Experimental: support for more than 90% of the latest ECMAScript specification. A sandbox that you can use to execute generated code with your #AI Coding Agent. #Development
#OpenSource MIT Lic

Local #GraphRAG framework with small model GLiNER-Relex trained specifically for building Graphs. Embedd #GraphDB or FalcorDB, #Neo4j, Memgraph
#OpenSource #LLM #AI
Apache 2.0
github.com/Knowledgator...

GitHub - tjjh89017/vrouter-operator: Deploy your vRouter(vyos) in your Kubernetes with KubeVirt Deploy your vRouter(vyos) in your Kubernetes with KubeVirt - tjjh89017/vrouter-operator

⭐️GitHub github.com/tjjh89017/vr...

vRouter Operator Demo: Manage VyOS Routers on Proxmox & KubeVirt with Kubernetes YouTube video by 伊達

Deploy your @vyos.bsky.social router using Infrastructure as a Code technique in your #Kubernetes as a VM. Update your config #GitOps-way.

GitHub link in the first 💬 👇
#OpenSource #k8s #Kubernetes #IaC
Apache 2.0 lic
Video: www.youtube.com/watch?v=Rsie...

GitHub - rtk-ai/rtk: CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies - rtk-ai/rtk

rtk — AI coding agent is drowning in CLI noise. RTK compresses outputs by 89%. #OpenSource, Rust.
MIT Lic
github.com/rtk-ai/rtk

HEDL - a #TOON alternative (compress JSON) for #LLM token savings and improved accuracy.
CLI, Python/JS bindings. Convert to & from #JSON. Also converts XML/YAML/CSV. Has LSP for VSCode.
Apache 2.0 lic #OpenSource #AI
github.com/dweve-ai/hedl

Work with #Gmail, Sheets, Drive & Docs via #CLI or #MCP
$ npm install -g @googleworkspace/cli
$ npx skills add github:googleworkspace/cli
#OpenSource #GoogleWorkspace
Apache 2.0 lic

github.com/googleworksp...

We're excited to announce ArcadeDB 26.3.1! Major news: redesigned Studio with an AI assistant, built-in MCP server support, geospatial indexing, a new TimeSeries data model, materialized views, hash indexes, 72 built-in graph algorithms, parallel query execution github.com/ArcadeData/a...

GitHub - microsoft/apm: Agent Package Manager Agent Package Manager. Contribute to microsoft/apm development by creating an account on GitHub.

#VibeCode Easy: declare and share #skills, prompts, instructions, and tools for your git project with APM.yaml across the team. By @microsoft.com #AI
Integrates with: #GitHub #Copilot, #Cursor, #Claude Code, #Codex, #Gemini
MIT lic #OpenSource #Development #DevOps
github.com/microsoft/apm

Release 26.2.1 · ArcadeData/arcadedb @ -0,0 +1,189 @@ ArcadeDB 26.2.1 Release Notes We're excited to announce ArcadeDB 26.2.1, a massive release that brings 484 commits and 200+ closed issues with significant advancements across the p...

ArcadeDB 26.2.1 is out! Biggest highlights:

- @Neo4j Bolt protocol support - use standard Neo4j drivers
- Full @openCypher TCK compliance
- New SQL parser
- SQL Triggers (SQL, JS, Java)
- Scheduled Backups

200+ issues closed
github.com/ArcadeData/a...

#graphdb #multimodel

GitHub - github/gh-aw-firewall Contribute to github/gh-aw-firewall development by creating an account on GitHub.

gh-aw #firewall - Agent network isolation by @github.com - execute #agents locally while creating a network isolation that prevents the agent to access certain websites. Perfect to prevent @openclaw-x.bsky.social from accessing restricted websites.
MIT lic
#AI #LLM #Networking #OpenSource

GitHub - temporalio/helm-charts: Temporal Helm charts Temporal Helm charts. Contribute to temporalio/helm-charts development by creating an account on GitHub.

@temporal.io is a distributed, scalable, durable, and highly available orchestration engine designed to execute asynchronous long-running business logic in a resilient way - perfect for #AI agents
#k8s #OpenSource
MIT license
github.com/temporalio/h...

GitHub - morph-kgc/morph-kgc: Powerful RDF Knowledge Graph Generation with RML Mappings Powerful RDF Knowledge Graph Generation with RML Mappings - morph-kgc/morph-kgc

morph-kgc to map #SQL, CSV, XML, JSON into #RDF triples to make a virtual #knowledgegraph.
Apache 2.0 lic.
#GraphRAG #RAG #Graph #OpenSource

github.com/morph-kgc/mo...

Implement the vchord / pgvector support by qdrddr · Pull Request #350 · vectorize-io/hindsight Implemented the vchord + pgvector support plan from code_update_plan.md. Here are the changes made: Added _detect_vector_extension() helper function in hindsight-api/hindsight_api/migrations.py ...

Just get my PR merged into amazing #Agentic Memory framework #Hindsight by @vectorize.bsky.social
Thanks for help @nicoloboschi.bsky.social!
Now you can benefit from VectorChord’s PG extension with up to 65k dimensions & BM25 Full Text Search. Enjoy!
#AI #Agents #RAG #LLM
github.com/vectorize-io...

FastMCP 3.0 is here!

This is our largest and most ambitious release ever.

We've re-architected the framework to support the next generation of MCP applications.

It's time to move fast and make things.

🧵

Observational Memory: 95% on LongMemEval - Mastra Research How we built a human-inspired memory system that achieves the highest score ever recorded on LongMemEval

🔗 mastra.ai/research/obs...
⭐️ GitHub github.com/mastra-ai/ma...