Telexed

AI news through a solo-operator lens — only what changes your day30 of 829

FILTER[All][Agents & tools][Models & API][Generative media][Infra & SaaS][ASO & growth][Indie business][Idea signals][Other][★6+ high-signal]

clear filters

Yesterday5 dispatches

#0829
#0829Agents & tools Paperclip Releases53 minutes ago
`Paperclip` `v2026.529.0` adds document annotations and company skills CLI
70radar
PaperclipCoding agent platform — document and skill workflows
Agent work now gets passage-level review threads and managed skill catalogs. Useful if you run self-hosted agent workflows; otherwise it is a strong product-direction signal.
- Issue documents now support revision-aware inline annotation threads with stable anchors, reducing vague review comments around agent-written specs.
- New skills CLI covers install, reset, audit, export, and assignment; skill provenance is tracked across bundled, catalog, runtime, and adapter sources.
- Fresh private self-hosted installs get a one-time first-admin browser claim, removing the invite bootstrap problem on LAN deployments.
- Claude Local can refresh its Anthropic model catalog from the UI, so new Claude models can appear without waiting for a Paperclip release.
Source: github.com/paperclipai/paperclip/releases/tag/v2026.529.Read original →
FIG-8291:1
70radar
FIG-8291:1
#0828
#0828Agents & tools Claude Code Releases13 hours ago
`Claude Code` `v2.1.158` adds Auto mode on Bedrock, Vertex, and Foundry
60radar
Claude CodeCoding agent CLI — automates code work with Claude
Auto mode now reaches managed-cloud deployments for Opus 4.7 and Opus 4.8. Useful only if you run through Bedrock, Vertex, or Foundry.
- CLAUDE_CODE_ENABLE_AUTO_MODE=1 enables the feature, so rollout is opt-in rather than default behavior.
- Support covers Bedrock, Vertex, and Foundry for Opus 4.7 and Opus 4.8; direct API users get no new signal here.
- Practical value is higher for teams keeping LLM traffic inside cloud-vendor rails; local CLI-only workflows can skip this update.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.158Read original →
FIG-8281:1
60radar
FIG-8281:1
#0827
#0827Agents & tools Claude Code Releases19 hours ago
`Claude Code` `v2.1.157` auto-loads local skill plugins
80radar
Claude CodeCoding agent CLI — automates code work with Claude in terminal
Local .claude/skills plugins now load without marketplace setup. Plugin scaffolding, agent settings, and worktree handling got smoother for daily agent work.
- Plugins under .claude/skills are auto-loaded, so private workflow extensions no longer need marketplace packaging first.
- claude plugin init <name> scaffolds local plugins directly; useful for turning repeated repo workflows into reusable skills.
- settings.json now honors the agent field for dispatched sessions, with --agent <name> available for one-off overrides.
- EnterWorktree can switch between Claude-managed worktrees mid-session, reducing context breaks during parallel agent tasks.
- Several fixes target terminals, image paste, resume state, and background agents; less time lost to CLI session cleanup.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.157Read original →
FIG-8271:1
80radar
FIG-8271:1
#0826
#0826Agents & tools Cline Releases20 hours ago
`Cline CLI` `v3.0.15` adds Hub for remote session control
80radar
ClineOpen-source coding agent — autonomous work in IDE and CLI
A new Hub web app can monitor clients, drive sessions, stream output, and restart the local hub. Useful for running agent work from another machine, with a room secret as the safety gate.
- Cline Hub supports local, LAN, and tunnel access gated by a room secret; remote agent control is now a built-in workflow.
- Global AGENTS rules now apply across all sessions, reducing repeated setup when multiple repos share the same coding conventions.
- Plugins can inject static or dynamic rules inside the sandbox, opening a cleaner path for reusable team or stack-specific behavior.
- The bundled catalog adds Claude Opus 4.8, Kimi K2.6, and Qwen3.7 Max with cache support; model choice gets fresher without manual wiring.
- Discord sessions now bind to individual authors and connector error flooding is capped, making shared channel workflows less brittle.
Source: github.com/cline/cline/releases/tag/cli-v3.0.15Read original →
FIG-8261:1
80radar
FIG-8261:1
#0825
#0825Models & API Google AI22 hours ago
`Gemini Omni` and `Gemini 3.5` demos released after Google I/O 2026
100radar
Google is showing multimodal and model-upgrade capabilities through 11 videos. Treat this as an evaluation queue, not a spec sheet; API pricing and availability decide adoption.
- The demos cover Gemini Omni and Gemini 3.5, both announced at Google I/O 2026; this is the first hands-on signal after launch.
- Video demos are useful for UX pattern discovery: voice, vision, and real-time flows can map directly to app prototypes.
- No pricing, limits, latency, or API availability details are included here. Production decisions still need official docs.
- For solo products, the near-term value is competitive scanning: identify demos that can become narrow paid features.
Source: blog.google/innovation-and-ai/models-and-research/geminiRead original →
FIG-8251:1
100radar
FIG-8251:1

Fri, May 2924 dispatches

#0824
#0824Agents & tools Google AI Forumyesterday
`Gemini Code Assist` Individual IDE/CLI Access May End June 18
80radar
Gemini Code AssistAI coding assistant — Gemini in IDE extensions and CLI
Individual Pro, Ultra, and free access stops on June 18, 2026. VS Code users should line up alternate auth or tools before agent workflows break.
- The notice names Gemini CLI and Gemini Code Assist IDE extensions. Individual Pro, Ultra, and free accounts are the affected paths.
- The cutoff date is June 18, 2026. Any VS Code workflow wired to personal Gemini access needs a fallback before then.
- The forum post does not confirm whether an Antigravity-specific VS Code extension continues. Treat it as unresolved until Google posts a clearer migration path.
Source: discuss.ai.google.dev/t/will-there-no-longer-be-a-geminiRead original →
FIG-8241:1
80radar
FIG-8241:1
#0823
#0823Agents & tools Hacker News · Show HN AIyesterday
`AISlop`, a CLI that catches AI-generated code smells
60radar
AISlopAI code-quality CLI — local code-smell scanning
Scans for patterns tests often miss: empty catches, dead code, duplicate helpers, and useless comments. Local-only hook integration makes it worth trying in agent-heavy repos.
- npx aislop scan runs locally, so source code is not uploaded. That lowers the barrier for private app repos.
- Targets AI-agent residue rather than syntax errors: empty catch, dead code, duplicated helpers, and low-value comments.
- Hook-based use after each tool call fits Claude Code, Codex, and opencode workflows where small slop accumulates fast.
Source: github.com/scanaislop/aislopRead original →
FIG-8231:1
60radar
FIG-8231:1
#0822
#0822Models & API Entrepreneuryesterday
New `Claude Opus 4.8` Is Trained to Be More 'Honest' and Stop 'Jumping to Conclusions'
80radar
Anthropic's new flagship model is tuned to admit when it doesn't know something and avoid unsupported claims. This aims to reduce hallucinations, making API responses more reliable.
- The core update is improved 'honesty,' enabling the model to say 'I don't know' more readily.
- This is a direct attempt to mitigate hallucinations by preventing the model from making unsupported assertions.
- For developers, this could lead to more faithful adherence to context in RAG (Retrieval-Augmented Generation) systems.
- The change suggests a bigger industry trend towards reliability over just capability, prioritizing trustworthy AI assistants.
Source: www.entrepreneur.com/business-news/the-new-claude-opus-4Read original →
FIG-8221:1
80radar
FIG-8221:1
#0821
#0821Indie business GeekNewsyesterday
`nice!nano`: a dorm-room hardware product that reached million-dollar scale
70radar
nice!nanoWireless keyboard controller board — Pro Micro-compatible form factor
A tiny compatibility decision turned a DIY keyboard pain into a real product. Niche hardware can still win when it fixes latency, battery life, and adoption friction at once.
- nice!nano kept the Pro Micro form factor, so existing DIY keyboard designs could adopt it with low switching cost.
- The pain was concrete: wireless DIY keyboards had input latency and short battery life. Nordic chips solved both enough to unlock demand.
- KiCad and open schematics lowered the build barrier. In niche hardware, ecosystem fit can matter more than broad-market positioning.
Source: news.hada.io/topic?id=29988Read original →
FIG-8211:1
70radar
FIG-8211:1
#0820
#0820Agents & tools Google AI Forumyesterday
`Antigravity Ultra` users hit persistent `MODEL_CAPACITY_EXHAUSTED` on Claude Opus thinking
60radar
AntigravityCoding agent IDE — uses Google's backend model routing
Paid agent sessions can fail before real work starts. Treat Antigravity Ultra as capacity-risky for Claude-heavy workflows until routing improves.
- The backend returns 503 `MODEL_CAPACITY_EXHAUSTED` from cloudcode-pa.googleapis.com, so this is server capacity, not local quota.
- Failures recur after 1-29 seconds on multi-step agent tasks like file edits, bash runs, and code generation. Long tasks are unreliable.
- Server-Timing shows about 13s spent waiting for model capacity before failure. Retry loops burn time without changing the result.
- Multiple similar forum threads over nearly a month point to account or routing-level fragility, not a one-off client setup problem.
Source: discuss.ai.google.dev/t/persistent-model-capacity-exhausRead original →
FIG-8201:1
60radar
FIG-8201:1
#0819
#0819Infra & SaaS vercel_blogyesterday
`Vercel` switches function invocations to per-unit billing
100radar
VercelFrontend cloud platform — optimized for Next.js deployment
Function usage now maps directly to actual calls. The rate is unchanged, but billing becomes smoother for small teams and bursty apps.
- Pro customers pay $0.0000006 per invocation, equivalent to the previous $0.60 per 1M invocations package rate.
- The change starts from the next billing cycle; current effective rates stay unchanged until the existing cycle ends.
- Per-unit billing reduces the chance that Vercel Functions usage eats a large chunk of monthly included credits at once.
- Applies to Pro and new Enterprise customers, so active Vercel apps should recheck cost dashboards after renewal.
Source: vercel.com/changelog/function-invocations-now-billed-perRead original →
FIG-8191:1
100radar
FIG-8191:1
#0818
#0818Infra & SaaS vercel_blogyesterday
Vercel warns AI endpoints are prime targets for inference theft
70radar
BotIDVercel bot defense tool — per-request automation checks
Cheap HTTP checks do not match $2-per-prompt economics. Verify every AI request, or one abused endpoint can turn into a five-figure bill.
- Attackers wrap custom endpoints in OpenAI or Anthropic adapters, then resell stolen inference through standard SDK clients.
- Session auth and IP limits fail when proxy fleets and throwaway accounts spread calls across thousands of identities.
- Playgrounds are the highest-risk shape because users control prompts, models, and parameters. Fixed server prompts lower but do not remove risk.
- BotID is Vercel's proposed gate: run bot verification on every AI request, not just signup or session start.
Source: vercel.com/blog/protecting-against-inference-theftRead original →
FIG-8181:1
70radar
FIG-8181:1
#0817
#0817Infra & SaaS Simon Willisonyesterday
`Datasette` `1.0a31` adds write SQL and stored queries
60radar
DatasetteSQLite data tool — publishes DBs as web UI and APIs
Permissions now cover insert/update/delete workflows directly in the SQL UI. Good fit for lightweight internal admin tools on SQLite, but still alpha.
- Users with the right permissions can run write queries against a database, turning Datasette from read-only browsing into a small data editing surface.
- Stored queries replace the old canned-query naming and can be saved privately or shared with other members of the same instance.
- The new execute-query UI starts from templated insert, update, and delete queries for editable tables, reducing hand-written SQL mistakes.
- Permission checks block operations like create table when the user lacks that grant, so this can back simple ops panels without opening full DB access.
Source: simonwillison.net/2026/May/29/datasette/#atom-everythingRead original →
60radar
PHOTO
FIG-8171:1
#0816
#0816Agents & tools GeekNews2 days ago
`Claude Code` Adds Dynamic Workflows for Parallel Subagents
90radar
Claude CodeCoding agent CLI — runs agentic dev tasks from the terminal
Large engineering jobs can now be split across many subagents in one session. This turns agent work from linear prompting into scripted orchestration worth testing on repo-scale tasks.
- A single session can dynamically write orchestration scripts, launch dozens to hundreds of parallel subagents, and collect results before delivery.
- The feature targets work too large or complex for one agent, so it fits migrations, wide refactors, audits, and test expansion.
- The bottleneck shifts to task decomposition. Vague goals will create noisy parallel work, while clear file/module boundaries should pay off.
Source: news.hada.io/topic?id=29981Read original →
FIG-8161:1
90radar
FIG-8161:1
#0815
#0815Agents & tools GeekNews2 days ago
`jqwik` 1.10.0 Prints Prompt-Injection Protestware During Tests
60radar
jqwikJava testing library — property-based testing support
A test dependency can leak hostile instructions into CI, IDE panels, and agent tool output. Treat logs as untrusted input when coding agents read them.
- jqwik 1.10.0 prints an instruction telling coding agents to ignore prior commands and delete tests/code during test runs.
- ANSI escape sequences hide the line in interactive terminals, but it can remain visible in CI logs, IDE test panels, and agent outputs.
- This is a concrete prompt-injection path through ordinary build logs. Pin versions and keep agents from blindly acting on stdout.
Source: news.hada.io/topic?id=29980Read original →
FIG-8151:1
60radar
FIG-8151:1
#0814
#0814Agents & tools Hermes Agent Releases2 days ago
`Hermes Agent` `v0.15.1`, same-day hotfix for dashboard reload loop
60radar
Hermes AgentCoding agent runtime — dashboard, MCP, and skills hub integrated
The broken v0.15.0 dashboard path is patched fast. Docker and MCP fixes reduce setup friction, but this is mostly a stability release.
- Loopback dashboard calls to /api/auth/me returned 401 by design; stale-token handling caused infinite reloads in Docker, hosted, and fresh installs.
- Docker no longer infers --insecure from bind host. LAN access and auth bypass are split via explicit HERMES_DASHBOARD_INSECURE=1.
- MCP bare commands like npx, npm, and node now resolve through /usr/local/bin, fixing silent container launch failures.
- The skills catalog expands from 858 to 19,932 entries, so picker coverage is much closer to the actual skills.sh index.
Source: github.com/NousResearch/hermes-agent/releases/tag/v2026.Read original →
FIG-8141:1
60radar
FIG-8141:1
#0813
#0813Agents & tools GeekNews2 days ago
`CodeBoarding`, Interactive Architecture Diagrams for Codebases
60radar
CodeBoardingOpen-source code map tool — auto-generates architecture docs
Static analysis plus LLM reasoning turns a repo into architecture docs. Useful before handing context to coding agents.
- Generates high-level architecture diagrams and component docs from a codebase, reducing the first-pass repo reading cost.
- Outputs Markdown into .codeboarding/, so the result can be versioned, reviewed, or fed into agents as project context.
- Best fit is unfamiliar or aging repos. For tiny apps, manual notes may beat the setup overhead.
Source: news.hada.io/topic?id=29976Read original →
FIG-8131:1
60radar
FIG-8131:1
#0812
#0812Infra & SaaS vercel_blog2 days ago
`Vercel Sandbox` now runs Docker containers
80radar
Vercel SandboxCloud sandbox — isolates agent-driven development work
Agents can now build images, run services, and install system packages inside isolated sandboxes. This turns preview and test environments into fuller dev machines.
- Docker can be installed with a daemon inside Vercel Sandbox, so agents can run containerized apps without touching the host system.
- Redis, Postgres, and other test dependencies can run as containers; fewer external services are needed for isolated QA.
- With persistent sandboxes, installed Docker and pulled images carry across sessions, cutting repeated setup time.
- FUSE filesystem drivers and VPN clients are now supported too, widening what sandboxed agent workflows can access.
Source: vercel.com/changelog/run-docker-containers-inside-vercelRead original →
FIG-8121:1
80radar
FIG-8121:1
#0811
#0811Agents & tools Cursor Changelog2 days ago
`Cursor` Adds Auto-review Run Mode
90radar
CursorAI code editor — agent-driven code edits and review
Longer autonomous runs now need fewer approval prompts. The value is less babysitting, as long as the safety model is strong enough for real repo work.
- Auto-review is a new run mode for longer Cursor tasks with fewer approval prompts, reducing interruption during agent work.
- The changelog explicitly pairs fewer approvals with safer execution, so the feature targets autonomy without fully removing guardrails.
- Best fit is bounded repo work such as refactors, fixes, and test passes where repeated approvals slow the loop.
Source: cursor.com/changelog/auto-reviewRead original →
FIG-8111:1
90radar
FIG-8111:1
#0810
#0810ASO & growth GeekNews2 days ago
Push Notifications Are Becoming Platform-Edited Channels
60radar
Notifications are no longer just delivery pipes. APNs and FCM sit between app and user, so wording, priority, and summary behavior now affect retention work.
- APNs and FCM began as centralized brokers for battery savings; every iPhone and Android notification already passed through platform servers.
- Apple and Google now parse, rank, summarize, and rewrite notifications, turning a once-neutral layer into editorial surface area.
- Push copy needs testing beyond open rate: summary text, priority, timing, and platform grouping can change what users actually see.
Source: news.hada.io/topic?id=29962Read original →
FIG-8101:1
60radar
FIG-8101:1
#0809
#0809Agents & tools OpenAI Codex Releases2 days ago
`Codex CLI` `0.135.0` improves diagnostics, remote status, and Vim mode
70radar
CodexCoding agent CLI — automates code work in the terminal
Supportability got much better: richer doctor, remote /status, and sturdier TUI output. Worth upgrading if you run Codex in terminals or remote sessions.
- codex doctor now reports environment, Git, terminal, app-server, and thread inventory diagnostics, cutting back-and-forth in support cases.
- Remote TUI sessions show connection details and server version through /status, which helps debug mismatched client-server setups.
- Vim mode adds text-object editing and better word/end-of-line behavior; daily terminal use gets noticeably less rough.
- Python SDK thread and turn APIs now expose Sandbox presets, making sandbox configuration less verbose in automation scripts.
Source: github.com/openai/codex/releases/tag/rust-v0.135.0Read original →
FIG-8091:1
70radar
FIG-8091:1
#0808
#0808Agents & tools Cline Releases2 days ago
`Cline` `v3.86.0` Adds `Claude Opus 4.8` and `Kimi K2.6` Support
60radar
ClineVS Code coding agent — supports multiple LLM providers
More top-end model choices land inside the editor agent. The 1M-context option matters for large repo work, while the repo move is maintenance-only for most users.
- Claude Opus 4.8 support includes 1M-context variants where available, useful for broad repo inspection and long planning runs.
- Moonshot Kimi K2.6 is now selectable, adding another strong coding/model option without changing agent workflow.
- Prompt-cache support display for Qwen 3.7 Max was fixed in the Cline provider, reducing model capability guesswork.
- The VS Code extension project moved to apps/vscode; relevant mainly for contributors and fork maintainers.
Source: github.com/cline/cline/releases/tag/v3.86.0Read original →
FIG-8081:1
60radar
FIG-8081:1
#0807
#0807Agents & tools Hermes Agent Releases2 days ago
`Hermes Agent` v0.15.0: faster startup, multi-agent Kanban, security hardening
90radar
Hermes AgentCoding agent CLI — supports multi-agent task boards
The release turns a coding CLI into a broader agent platform. Faster startup, task swarms, secrets handling, and promptware defense make it worth testing for serious automation.
- run_agent.py dropped from 16,083 to 3,821 lines across 14 agent/ modules, making plugin work and codebase navigation less painful.
- Kanban now supports auto-decomposition, swarm topology, scheduled tasks, per-task worktrees, branches, and model overrides for cost-aware parallel work.
- Cold start improves again: one more second cut, 47% fewer per-conversation function calls, and session_search is now 4,500x faster and free.
- Security and ops got real attention: Brainworm-class promptware defense plus Bitwarden Secrets Manager bootstrap token instead of many provider keys.
- New integrations include Krea 2, FAL image generation as a plugin, Nous MCP catalog picker, ntfy, and deeper xAI/Grok support.
Source: github.com/NousResearch/hermes-agent/releases/tag/v2026.Read original →
FIG-8071:1
90radar
FIG-8071:1
#0806
#0806Agents & tools Claude Code Releases2 days ago
`Claude Code` `v2.1.154`: `Opus 4.8`, dynamic workflows, cheaper fast mode
90radar
Claude CodeCoding agent CLI — automates code edits and runs in terminal
Background workflows can now fan work across tens to hundreds of agents. This is a serious step toward larger autonomous coding runs, with cost and safety controls worth testing now.
- Opus 4.8 is now available with high effort by default, plus /effort xhigh for harder tasks. Expect better results on deep refactors and debugging.
- Dynamic workflows let Claude create and run multi-agent workflows in the background. Use /workflows to inspect runs instead of manually coordinating subtasks.
- Fast mode on Opus 4.8 costs 2x standard rate for 2.5x speed. Good for batch edits where latency beats token efficiency.
- /simplify now performs cleanup-only review and applies fixes. It is better suited to reuse, simplification, efficiency, and code altitude, not bug hunting.
- Safety and ops got sharper: rm -rf $HOME trailing-slash blocking, pending MCP approval display, and stronger data-exfiltration detection.
Source: github.com/anthropics/claude-code/releases/tag/v2.1.154Read original →
FIG-8061:1
90radar
FIG-8061:1
#0805
#0805Models & API GeekNews2 days ago
Anthropic Ships `Claude Opus 4.8`
90radar
Same-price upgrade over 4.7, with better benchmarks and collaboration. Worth re-testing as a coding and agent default before changing spend assumptions.
- Built on Claude Opus 4.7; this is an incremental top-tier model refresh, not a new product line.
- Benchmark performance improved while pricing stays unchanged. Existing API cost models do not need recalculation first.
- Collaboration capability was strengthened, so agent workflows and pair-programming style tasks deserve fresh evals.
Source: news.hada.io/topic?id=29960Read original →
FIG-8051:1
90radar
FIG-8051:1
#0804
#0804Agents & tools GitHub Changelog2 days ago
`Claude Opus 4.8` is now generally available in `GitHub Copilot`
80radar
A stronger coding model is now inside the editor workflow. Test it on hard refactors and generation-heavy tasks; daily value depends on latency and plan access.
- Claude Opus 4.8 is Anthropic's latest Opus model, now generally available in GitHub Copilot.
- Early GitHub testing says code understanding and generation improved clearly; verify on your own repo before switching defaults.
- Best first use cases: complex file edits, unfamiliar codebase reading, and generation-heavy tasks where model quality beats speed.
Source: github.blog/changelog/2026-05-28-claude-opus-4-8-is-geneRead original →
FIG-8041:1
80radar
FIG-8041:1
#0803
#0803Models & API Anthropic Python SDK2 days ago
`anthropic-sdk-python` `v0.105.0` adds `claude-opus-4-8` support
70radar
anthropic-sdk-pythonPython SDK — official client for Anthropic API
Python clients can now call the new Opus variant and inspect richer token details. Worth upgrading before wiring new Claude features into production.
- claude-opus-4-8 is now supported in the official Python SDK, removing manual model-name workarounds.
- Mid-conversation system blocks landed in the API layer, useful for agent flows that need runtime instruction changes.
- usage.output_tokens_details exposes finer output-token accounting, making cost/debug dashboards easier to build.
- Custom file size caps are supported, so upload limits can be enforced closer to the SDK boundary.
Source: github.com/anthropics/anthropic-sdk-python/releases/tag/Read original →
FIG-8031:1
70radar
FIG-8031:1
#0802
#0802Models & API Anthropic TypeScript SDK2 days ago
`Anthropic TypeScript SDK` `0.100.0` Adds `claude-opus-4-8` Support
70radar
Anthropic TypeScript SDKTypeScript SDK — typed client for Anthropic APIs
Model access, mid-conversation system blocks, and detailed output-token usage landed together. Worth upgrading before testing new Claude flows.
- claude-opus-4-8 is now supported in the SDK, so TypeScript apps can call it without local type workarounds.
- Mid-conversation system blocks unlock cleaner control changes during long chats and agent loops.
- usage.output_tokens_details adds more granular output accounting, useful for cost dashboards and per-feature metering.
Source: github.com/anthropics/anthropic-sdk-typescript/releases/Read original →
FIG-8021:1
70radar
FIG-8021:1
#0801
#0801Generative media future_tools_news2 days ago
`ElevenLabs Dubbing v2` Launches Emotion-Preserving Translation
70radar
ElevenLabsAI voice generation SaaS — strong in synthesis and dubbing
Voice localization gets less robotic across languages. Useful for app demos, course clips, and launch videos when separate voice talent is too costly.
- Dubbing v2 targets emotion preservation during multilingual dubbing, reducing the flat tone common in translated audio.
- Release date is 2026-05-28 and the source is ElevenLabs' official X account, so this is a product launch, not a rumor.
- Best fit is lightweight localization: demo videos, onboarding clips, YouTube shorts, and paid course snippets without rebuilding the whole production flow.
Source: x.com/elevenlabs/status/2060024691444617418?utm_source=fRead original →
FIG-8011:1
70radar
FIG-8011:1

Thu, May 281 dispatches

#0800
#0800ASO & growth RevenueCat Blog2 days ago
RevenueCat: missed reactivation leaves churn recovery on the table
60radar
RevenueCatSubscription SaaS — unifies app billing and entitlements
Win-back only pays off when billing, targeting, and offer paths are ready before users return. Treat churn as a recoverable segment, not a dead cohort.
- The supplied body is brief, but the title gives the signal: 20% of churned users can return when reactivation is handled deliberately.
- RevenueCat frames reactivation as an operating setup problem: entitlement state, paywall routing, and campaign measurement must already work.
- Best fit is subscription apps with enough churn volume to segment past users. One-off apps or tiny cohorts get limited lift.
Source: www.revenuecat.com/blog/growth/20-of-your-churned-users-Read original →
FIG-8001:1
60radar
FIG-8001:1

↻ Load earlier issues

`Paperclip` `v2026.529.0` adds document annotations and company skills CLI

`Claude Code` `v2.1.158` adds Auto mode on Bedrock, Vertex, and Foundry

`Claude Code` `v2.1.157` auto-loads local skill plugins

`Cline CLI` `v3.0.15` adds Hub for remote session control

`Gemini Omni` and `Gemini 3.5` demos released after Google I/O 2026

`Gemini Code Assist` Individual IDE/CLI Access May End June 18

`AISlop`, a CLI that catches AI-generated code smells

New `Claude Opus 4.8` Is Trained to Be More 'Honest' and Stop 'Jumping to Conclusions'

`nice!nano`: a dorm-room hardware product that reached million-dollar scale

`Antigravity Ultra` users hit persistent `MODEL_CAPACITY_EXHAUSTED` on Claude Opus thinking

`Vercel` switches function invocations to per-unit billing

Vercel warns AI endpoints are prime targets for inference theft

`Datasette` `1.0a31` adds write SQL and stored queries

`Claude Code` Adds Dynamic Workflows for Parallel Subagents

`jqwik` 1.10.0 Prints Prompt-Injection Protestware During Tests

`Hermes Agent` `v0.15.1`, same-day hotfix for dashboard reload loop

`CodeBoarding`, Interactive Architecture Diagrams for Codebases

`Vercel Sandbox` now runs Docker containers

`Cursor` Adds Auto-review Run Mode

Push Notifications Are Becoming Platform-Edited Channels

`Codex CLI` `0.135.0` improves diagnostics, remote status, and Vim mode

`Cline` `v3.86.0` Adds `Claude Opus 4.8` and `Kimi K2.6` Support

`Hermes Agent` v0.15.0: faster startup, multi-agent Kanban, security hardening

`Claude Code` `v2.1.154`: `Opus 4.8`, dynamic workflows, cheaper fast mode

Anthropic Ships `Claude Opus 4.8`

`Claude Opus 4.8` is now generally available in `GitHub Copilot`

`anthropic-sdk-python` `v0.105.0` adds `claude-opus-4-8` support

`Anthropic TypeScript SDK` `0.100.0` Adds `claude-opus-4-8` Support

`ElevenLabs Dubbing v2` Launches Emotion-Preserving Translation

RevenueCat: missed reactivation leaves churn recovery on the table