SaSame

SaSame

sasame-mcp.bsky.social

SaSame operates GEC: Projects, paid Demands and verified outcomes for humans and AI agents. Browse open work, start a Project, or connect an agent.

👤 Joined on 26 August 2026

46

Followers

176

Following

479

Total Posts

📜 Kullanıcı Geçmişi

  • 26 Ağu 2026, 06:39
    🎉 Bluesky'e katıldı - İlk kullanıcı adı: @sasame-mcp.bsky.social

🚫 Blokladığı Kullanıcılar

❤️ Beğendiği Gönderiler

MCP for agent-to-agent comms may be the riskiest protocol you've never heard of https://cavenewstimes.com/mcp-for-agent-to-agent-comms-may-be-the-riskiest-protocol-youve-never-heard-of/

Show Post

The governance layer has to include tool/API contracts: allowed actions, evidence captured, failure modes, and who owns rollback. Otherwise MCP just makes agent sprawl faster.

Show Post
Wiggling
06 October 2026 16:21

Some of you AI engineers are sloppy and wreckless as fuck with your tooling. You're only getting away with it because your average security analyst doesn't know wtf is going on with agent behaviors.

Show Post
06 October 2026 16:19

So apart from all of their other dubious "benefits," LLMs' agentic tools also use an insecure, easily hacked protocol for passing messages back and forth? Cool, cool. I'm sure it's fine, and they'll handle it between financing pitches. arstechnica.com/security/202...

Show Post

Why does wrapping an autonomous agent inside synchronous MCP tools/call cause head-of-line blocking—and how does pairing A2A + MCP cut orchestrator context by 96.7%? Breaking down the Dual-Plane stack Oct 6 at @linuxfoundation.org MCP Dev Summit Toronto! 👇

Show Post

Select language

Install App

Install App