#endpoints
implemented a couple more endpoints, it's not broken right RIGHT
November 11, 2026 at 5:03 AM
Most of TOR’s endpoints are owned by the Feds now 😭
November 10, 2026 at 9:14 PM
Arguably both endpoints should be no lol
November 10, 2026 at 10:12 PM
it just needs to reach out via curl or whatever to any endpoints i think up
November 10, 2026 at 2:38 AM
i will never purchase a roku device ever again
November 10, 2026 at 4:20 PM
udp-replicator (⭐️ 59)

A tiny UDP proxy that can replicate traffic to one or more endpoints.

#go
November 11, 2026 at 7:55 AM
DT120, an investigational orally disintegrating tablet formulation of LSD, met the primary endpoints in a pair of phase III trials for generalized anxiety disorder and major depressive disorder. @dankarlin.com https://www.medpagetoday.com/meetingcoverage/neicongress/123772
Oral LSD Drug Hits Phase III Endpoints in Anxiety, Depression
Voyage and Emerge patients experienced significant reductions in symptoms within a week
www.medpagetoday.com
November 10, 2026 at 1:36 AM
3/ Compare all cloud GPU & serverless endpoints hosting GPT-4o Mini (<400ms TTFT SLA) on our live calculator:
https://tokenmargin.dev/models/gpt-4o-mini-pricing-calculator.html?ref=bsky #AI #LLM #FinOps #DevTools #BuildInPublic
November 11, 2026 at 1:45 PM
i still suspect a malware could hit the endpoints that the MCP server thing exposes but also: if a malware runs on your computer "can it change the settings for my GPU" is probably the least of your concerns lmao
November 8, 2026 at 10:39 AM
3/ Compare all cloud GPU & serverless endpoints hosting Claude 3.7 Sonnet (<400ms TTFT SLA) on our live calculator:
https://tokenmargin.dev/models/claude-3-7-sonnet-pricing-calculator.html?ref=bsky #AI #LLM #FinOps #DevTools #BuildInPublic
November 11, 2026 at 3:15 PM
3/ Compare all cloud GPU & serverless endpoints hosting GPT-4o Mini (<400ms TTFT SLA) on our live calculator:
https://tokenmargin.dev/models/gpt-4o-mini-pricing-calculator.html?ref=bsky #AI #LLM #FinOps #DevTools #BuildInPublic
November 10, 2026 at 6:13 PM
3/ Compare all cloud GPU & serverless endpoints hosting Claude 3.7 Sonnet (<400ms TTFT SLA) on our live calculator:
https://tokenmargin.dev/models/claude-3-7-sonnet-pricing-calculator.html?ref=bsky #AI #LLM #FinOps #DevTools #BuildInPublic
November 10, 2026 at 5:45 PM
what's the reason for going through `agy` instead of one of the gemini api endpoints?
November 6, 2026 at 1:51 AM
A server’s entry point is a single file in your project.

Write it as a request handler with the same signature as a function, or bring an Express, Hono, or Fastify app as it is. No ports to configure, no health check endpoints, no adapters.
November 10, 2026 at 7:31 PM
Turns out it blocks some of my VPN endpoints. Fun.
November 7, 2026 at 9:05 PM
would it be possible to unblock download endpoints for release artifacts so apps like obtainium can pull them :plead:
November 7, 2026 at 7:32 AM
I wonder how far along 12.2 is, usually we'd have seen it hit the encrypted endpoints at this point. 12.1.7 is ~4 weeks out if they keep to the normal cadence and I doubt they want to put up a PTR around Christmas if they push the cadence by 1-2 weeks, but who knows.
November 9, 2026 at 7:59 PM
Rapid prototyping tools mean security guardrails can easily be skipped. How does your team audit public endpoints and authentication when building with AI agents?
November 11, 2026 at 6:50 PM
Wrapping a REST API as an MCP server by hand means guessing which endpoints are safe for an agent to call. MCP Studio, from, turns an OpenAPI 3.0/3.1 spec into a reviewed MCP server project. Endpoints get read, write or destructive labels. It's free, and exports config for Claude Code, Cursor,...
November 10, 2026 at 3:12 AM
3/ Compare all cloud GPU & serverless endpoints hosting DeepSeek R1 (671B Reasoning) (<400ms TTFT SLA) on our live calculator:
https://tokenmargin.dev/models/deepseek-r1-pricing-calculator.html?ref=bsky #AI #LLM #FinOps #DevTools #BuildInPublic
November 11, 2026 at 1:45 AM
Functions are built on one idea: each invocation handles one request. That’s what makes them easy to scale.

It gets in the way once a backend outgrows a set of endpoints: holding a WebSocket open, waiting on an AI model for long periods of time, keeping state between requests.
November 10, 2026 at 7:31 PM
Service Connect endpoints that access global Google APIs. For more information, see Internal ranges for Private Service Connect endpoints.
November 9, 2026 at 9:20 PM
ML-007C-MA, an investigational muscarinic M1/M4 receptor agonist that combines betovumeline and fesoterodine fumarate, significantly improved schizophrenia symptoms in adults who had an acute exacerbation of psychosis, data showed.
https://www.medpagetoday.com/meetingcoverage/neicongress/123795
M1/M4 Agonist Relieved Schizophrenia Symptoms in Mid-Stage Trial
Twice-daily dose hit primary and secondary endpoints
www.medpagetoday.com
November 11, 2026 at 12:13 AM
3/ Compare all cloud GPU & serverless endpoints hosting DeepSeek V3 (0324) (<400ms TTFT SLA) on our live calculator:
https://tokenmargin.dev/models/deepseek-chat-v3-0324-pricing-calculator.html?ref=bsky #AI #LLM #FinOps #DevTools #BuildInPublic
November 10, 2026 at 6:46 AM
The contingient factor there would be it being "past use", which it isn't?

They themselves rated it by a vibe coding metric, and on the rating scale of 0-7 with 0 being pure vibeslop and 7 being purely human made, they self-rated it a 1
github.com/dtupper/ntfs...
GitHub - dtupper/ntfs-meta: Minimal read-only NTFS metadata server (MFT + USN journal) with REST and MCP endpoints, for local disk-space agents
Minimal read-only NTFS metadata server (MFT + USN journal) with REST and MCP endpoints, for local disk-space agents - dtupper/ntfs-meta
github.com
November 10, 2026 at 1:57 AM