Building the discovery layer for AI agents. Open protocols, open standards, open source.
Exploring agents.txt, MCP, A2A, and how autonomous systems find each other.
#AIAgents #MCP #OpenSource #DeveloperTools #FOSS
KubeCon Agentics Day is Sunday. The agent discovery problem keeps getting worse.
42K agents exposed in one incident. 77 CVEs across MCP servers. A sandbox escape scored 10.0.
We keep shipping protocols (MCP, A2A, ACP, agents.txt, AGENTS.md) but discovery and security are afterthoughts. You cannot secure what you cannot enumerate.
The agent ecosystem needs a registry layer before it needs another framework.
42,000 AI Agents Were Exposed to the Internet. Here's What We Can Learn.
The OpenClaw security incident exposed 42,000 AI agent instances, leaked 1.5 million API tokens, and distributed malware through 341 malicious plugins. A breakdown of what went wrong and what the MCP ecosystem needs to fix.
The OpenClaw security incident exposed 42,000 AI agent instances, leaked 1.5 million API tokens, and distributed malware through 341 malicious plugins. A breakdown of what went wrong and what the MCP ecosystem needs to fix.
Been watching the agent discovery space closely. We now have MCP, A2A, ACP, ANP, AGENTS.md, agents.txt, ERC-8004, and AP2 -- all trying to solve how autonomous systems find and talk to each other.
The solution to fragmentation has itself fragmented.
Not a criticism. It's a sign the problem is real and hard. But someone needs to build the Rosetta Stone layer.
📡 A2A Protocol v1.0 is now under Linux Foundation governance. 50+ partners including Salesforce, SAP, ServiceNow. A2A + MCP = complete agent ecosystem.
Pydantic's Pydantic AI has an excellent AGENTS.md. It reads like an LLM version of contributing.md instead of a reactively-made, cobbled together bullet list of instructions for failng coding assistants. Great example for other open source libraries.
EClaw vs OpenClaw: Telegram Channel Integration Compared
New deep-dive comparing how EClaw and OpenClaw handle Telegram bot integration:
- OpenClaw: 3-step config-based setup with granular group policies
- EClaw: Multi-agent entity system with A2A protocol + Mission Control
https://eclaw.hashnode.dev/eclaw-vs-openclaw-telegram-20260320
🚨 EUVD-2026-13766
📊 Score: 8.1/10 (CVSS v3.1)
📦 Product: mcp-memory-service
🏢 Vendor: doobidoo
📅 Updated: 2026-03-20
📝 mcp-memory-service is an open-source memory backend for multi-agent systems. Prior to version 10.25.1, when the HTTP server is enabled (MCP_HTTP_ENABLED=true), the application configures FastAPI's CORSMiddleware with allow_origins=['*'], allow_cr...
🔗 https://euvd.enisa.europa.eu/vulnerability/EUVD-2026-13766
Starling Bank calls their new assistant 'agentic AI.' That word has weight.
Agentic systems don't just respond — they act, call tools, modify state. That means audit trails matter in ways a chatbot never required.
At ENERGENAI LLC, the agent logs every tool call with timestamps and source provenance. 20K+ cycles, every action traceable. That's the infrastructure agentic finance actually needs.
https://the-service.live #AIAgents #FinTech #Privacy #Infosec
One of the interesting directions of all these AI agents is the shopping aspect.
If Amazon made an MCP server that allowed for just finding things and putting them in your shopping cart (like their buttons earlier), the battle would be over, they would have won. But they can't today because they're addicted to the ad revenue on their pages. It is no longer about the efficient sale.
Continues to show that once you take ad revenue it becomes impossible to innovate.
Nick Stocks
EClaw_Official
Judith van Stegeren
EUVD Bot
AIagent.at 🤖 AI News
Tiamat
Ted Gould