tomjwxf

5 ms·

1. 1.▲
   Token spend is not token capital(twitter.com)

   3 points by tomjwxf 3mo ago|0 comments
2. 2.▲
   Token Capital Needs Receipts(twitter.com)

   2 points by tomjwxf 3mo ago|0 comments
3. 3.▲
   by tomjwxf 5mo ago

   Re classifier routing: text-shape signals (token count, syntactic markers) underspecify the boundary, especially for agent-generated queries. The signal that worked better in our policy-gated tool-call setting was the surrounding intent con
4. 4.▲
   by tomjwxf 6mo ago

   my core thesis is that AGI is here, it just needs accountability and efficient frameworks to navigate our arbitrary world
5. 5.▲
   by tomjwxf 6mo ago

   The gateway approach (OAuth + RBAC) solves the perimeter problem — who can connect. protect-mcp solves a different layer — what can they do once connected, and how do you prove it. It wraps any MCP server as a stdio proxy. Per-tool policies
6. 6.▲
   by tomjwxf 6mo ago

   The staged autonomy pattern ("trust is earnable") maps directly to what we built with protect-mcp — shadow mode first (log everything, block nothing), then enforce when you've seen enough data to trust the policies. For the p
7. 7.▲
   by tomjwxf 6mo ago

   This is exactly right. We implemented delegation receipts — Agent A grants scoped authority to Agent B, producing a signed receipt. B's subsequent actions reference A's delegation receipt. An auditor can trace the full chain from
8. 8.▲
   Signed receipts for MCP tool calls – prove what your agent did(npmjs.com)

   2 points by tomjwxf 6mo ago|0 comments