Не попадитесь на накрученные каналы! Узнайте, не накручивает ли канал просмотры или
подписчиков
Проверить канал на накрутку
Телеграм канал «Trend Rangers 🤖»
Trend Rangers 🤖
1.3K
11
4
1
0
Каждую неделю мы говорим о трендах в Tech:
http://trendrangers.tilda.ws/
http://trendrangers.tilda.ws/
Подписчики
Всего
56
Сегодня
0
Просмотров на пост
Всего
5
ER
Общий
8.28%
Суточный
7.14%
Динамика публикаций
Telemetr - сервис глубокой аналитики
телеграм-каналов
телеграм-каналов
Получите подробную информацию о каждом канале
Отберите самые эффективные каналы для
рекламных размещений, по приросту подписчиков,
ER, количеству просмотров на пост и другим метрикам
рекламных размещений, по приросту подписчиков,
ER, количеству просмотров на пост и другим метрикам
Анализируйте рекламные посты
и креативы
и креативы
Узнайте какие посты лучше сработали,
а какие хуже, даже если их давно удалили
а какие хуже, даже если их давно удалили
Оценивайте эффективность тематики и контента
Узнайте, какую тематику лучше не рекламировать
на канале, а какая зайдет на ура
на канале, а какая зайдет на ура
Показано 7 из 1 316 постов
Смотреть все посты
Пост от 27.10.2026 11:45
1
0
0
self.md Radar — 27 October 2026
Screenpipe makes waiting for AI redaction opt-in. A compaction study traces forged instructions into summaries. GPT-6.1 Terra lists $1 input and $5 output per million tokens.
Three stories, original sources and limits:
https://self.md/signals/2026-10-27-redaction-does-not-have-to-finish-first/
Screenpipe makes waiting for AI redaction opt-in. A compaction study traces forged instructions into summaries. GPT-6.1 Terra lists $1 input and $5 output per million tokens.
Three stories, original sources and limits:
https://self.md/signals/2026-10-27-redaction-does-not-have-to-finish-first/
Пост от 26.10.2026 11:52
1
0
0
self.md Radar — 26 October 2026
LangChain 1.7 changes tool-argument streaming. PaperFormalizer leaves main-result edits to authors by default. A retry study keeps unknown outcomes unresolved. Seam Cutter removes a strip from inside an image.
Four stories, original sources and caveats:
https://self.md/signals/2026-10-26-129-mb-for-one-file-write-call/
LangChain 1.7 changes tool-argument streaming. PaperFormalizer leaves main-result edits to authors by default. A retry study keeps unknown outcomes unresolved. Seam Cutter removes a strip from inside an image.
Four stories, original sources and caveats:
https://self.md/signals/2026-10-26-129-mb-for-one-file-write-call/
Пост от 25.10.2026 11:44
1
0
0
Google’s policy says a user choice can make Gmail, Drive and Photos data available to ad-supporting services. OneMemory makes public exposure opt-in; Octobench shows how the harness changes a model result.
Today’s Radar: three primary-source records about permission, network boundaries and coding-agent evaluation.
https://self.md/signals/2026-10-25-the-permission-the-port-the-branch-the-harness/
Today’s Radar: three primary-source records about permission, network boundaries and coding-agent evaluation.
https://self.md/signals/2026-10-25-the-permission-the-port-the-branch-the-harness/
Пост от 24.10.2026 11:10
1
0
0
Cloudflare says audit candidates should face a fresh verifier; task-view compares an agent’s map against Git; Kimi reports a cyber-safety trade-off.
Today’s Radar: three primary-source records about claims that still need a receipt.
https://self.md/signals/2026-10-24-a-place-for-the-unresolved-file/
Today’s Radar: three primary-source records about claims that still need a receipt.
https://self.md/signals/2026-10-24-a-place-for-the-unresolved-file/
Пост от 23.10.2026 22:45
1
0
0
on the record: “Your agent only sees that access is granted and doesn’t see your password.”
1Password’s agent-login flow keeps the credential in the vault while the browser receives a usable session. That is a narrow but useful boundary.
The remainder is still an account-permission problem: what can happen after the page is signed in?
Radar: https://self.md/signals/2026-10-23-a-browser-session-without-the-password/
Source: 1Password for Agents — retrieved 23 October 2026
1Password’s agent-login flow keeps the credential in the vault while the browser receives a usable session. That is a narrow but useful boundary.
The remainder is still an account-permission problem: what can happen after the page is signed in?
Radar: https://self.md/signals/2026-10-23-a-browser-session-without-the-password/
Source: 1Password for Agents — retrieved 23 October 2026
Пост от 23.10.2026 11:06
1
0
0
1Password lets an agent sign in without seeing your password.
Today’s Radar: credential custody, sandbox metadata allowlists, and the cost of repeated bad API keys.
https://self.md/signals/2026-10-23-a-browser-session-without-the-password/
Today’s Radar: credential custody, sandbox metadata allowlists, and the cost of repeated bad API keys.
https://self.md/signals/2026-10-23-a-browser-session-without-the-password/
Пост от 22.10.2026 10:47
1
0
0
kiso 0.51.0 can reject a checkpoint when a host-required H2 is missing. Graphify 0.9.93 redacts credential-shaped rationale labels before graph output. omnarad 0.2.9 stops duplicating and escaping tool results sent to the model.
Today’s source-checked Radar: https://self.md/signals/2026-10-22-a-header-can-be-a-requirement/
Today’s source-checked Radar: https://self.md/signals/2026-10-22-a-header-can-be-a-requirement/
