cuppa

brewed 03:03 AM
← Jun 15

Tuesday

jun16

2026

Jun 17 →

the brief

Platforms stretched their legs while agents multiplied. Vercel lifted Functions to 30‑minute runs and Next.js canaried Rspack 2.0, as Anthropic’s Claude Code, ByteDance’s Deer‑Flow, Nous’s Hermes Agent, and Datasette’s agent added sharper controls and workflows. A neat Cloudflare WAF trick and Axios’s Anthropic backstory rounded out a pragmatic, tool‑heavy day.

the poursit · sip · 8 items

pulse

(06)
  • vercel/news· feedJun 15, 02:00 PM

    Vercel Functions extended to 30 minutes

    Node.js and Python Functions on Pro/Enterprise can now run up to 30 minutes, enabling longer jobs with Fluid Compute pausing billing when idle; support for more runtimes is coming.

    Vercel Functions can now run up to 30 minutes — Vercel Functions using the Node.js and Python runtimes now support execution durations up to 30 minutes for Pro and Enterprise teams, more than 2x the previous 800 second limit. Support for additional runtimes is coming soon. Use longer-running Functions for work that needs more time to finish, including: Fluid Compute keeps long-running work cost-efficient. Active CPU billing only applies while your code is executing, and pauses while your Func...

    signal 8hype 1platform_updateserverlessvercel_functionssource ↗
  • vercel/next.js· feedJun 16, 12:16 AM

    Next.js canary updates to Rspack 2.0

    Canary 16.3.0 upgrades to Rspack 2.0 and tweaks docs and agent-related tooling, a meaningful build change Next.js apps should test against before the next stable.

    v16.3.0-canary.52 — Core Changes chore: upgrade next-rspack to rspack 2.0: #92222 Misc Changes docs: Route Handlers to be supported later on: #94705 [ci] Allow "rerun failed jobs" to work up until 5 days for e2e deploy tests: #94812 Fix stale next data API test URL: #94815 Format submillisecond durations as milliseconds: #94813 Instruct agents to keep and commit the generated agent-rules block: #94725 [agents-md] Index bundled docs instead of downloading into .next-docs: #94719 docs: add pref...

  • anthropics/claude-code· feedJun 15, 09:35 PM

    Claude Code adds parameter-aware permissions

    New Tool(param:value) permission syntax and nested .claude precedence give tighter control over subagents and skills, improving safety and determinism in multi-directory agent projects.

    v2.1.178 — What's changed Added Tool(param:value) syntax for permission rules to match a tool's input parameters (with * wildcard), e.g. Agent(model:opus) to block Opus subagents Skills in nested .claude/skills directories now load when working on files there; on a name clash, the nested skill appears as <dir>:<name> so both stay available Nested .claude/ directories: the agent, workflow, and output-style closest to the working directory now wins when names collide; project-scope workflow sav...

    signal 9hype 1release_notesclaude_codepermissionssource ↗
  • pplx/oss-rising-24h· researchJun 16, 03:00 AM

    ByteDance open-sources long-horizon SuperAgent

    Deer-Flow orchestrates sandboxes, memory, tools, skills, subagents, and a message gateway to tackle multi-hour tasks, offering a composable open-source harness for complex agent workflows.

    GitHub - bytedance/deer-flow: An open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of tasks that could take minutes to hours. — An open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of tasks that could take minutes to hours. - bytedance/deer-flow

    signal 8hype 1agent_frameworkopen_sourcegithub_reposource ↗
  • pplx/oss-rising-24h· researchJun 16, 03:00 AM

    Nous Research releases Hermes Agent

    Hermes Agent positions as an extensible, personalizable open-source agent framework that “grows with you,” inviting community contributions for building taskable assistants.

    GitHub - NousResearch/hermes-agent: The agent that grows with you — The agent that grows with you. Contribute to NousResearch/hermes-agent development by creating an account on GitHub.

    signal 6hype 2agent_frameworkopen_sourcegithub_reposource ↗
  • simonw/blog· feedJun 15, 05:19 PM

    datasette-agent adds approved write SQL tool

    Datasette-agent 0.3a0 adds execute_write_sql, requiring explicit user approval before database writes, bringing permissioned modifications and safer automation to Datasette’s agent workflows.

    datasette-agent 0.3a0 — <p><strong>Release:</strong> <a href="https://github.com/datasette/datasette-agent/releases/tag/0.3a0">datasette-agent 0.3a0</a></p> <blockquote> <ul> <li>New tool, <code>execute_write_sql</code>, which requests user approval and then writes to a database - taking user permissions into account. <a href="https://github.com/datasette/datasette-agent/issues/27">#27</a></li> </ul> </blockquote> <p>I added a mechanism for asking user approval in <a href="https://simonwillis...

    signal 8hype 1release_notesagent_toolspermissionssource ↗

findings

(01)

voices

(01)
  • simonw/blog· feedJun 15, 02:57 PM

    Axios recounts Anthropic outage tensions

    Willison highlights a sourced Axios account of tensions around the White House 'Fable' demo and Anthropic downtime—useful context on reliability, politics, and vendor risk in AI deployments.

    "They screwed us": Personality clashes sent Anthropic's models offline — <p><strong><a href="https://www.axios.com/2026/06/15/anthropic-white-house-fable-mythos">&quot;They screwed us&quot;: Personality clashes sent Anthropic&#x27;s models offline</a></strong></p> Lots of "source familiar with the administration's thinking" and "source close to Anthropic" in this Axios piece, which is the best collection of behind-the-scenes gossip I've seen about the US government <a href="https://simonwilli...

    signal 5hype 2anthropicoutagemedia_reportsource ↗