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

Fri, May 294 dispatches

#0829
#0829Infra & 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-8291:1
100radar
FIG-8291:1
#0828
#0828Infra & SaaS vercel_blog2 days ago
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-8281:1
70radar
FIG-8281:1
#0827
#0827Infra & SaaS Simon Willison2 days ago
`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-8271:1
#0826
#0826Infra & 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-8261:1
80radar
FIG-8261:1

Thu, May 283 dispatches

#0825
#0825Infra & SaaS vercel_blog3 days ago
`Amazon OpenSearch Serverless` lands in the `Vercel Marketplace`
70radar
Amazon OpenSearch ServerlessServerless search service — unified vector and keyword search
Search infra can now be provisioned from the Vercel dashboard. Built-in env injection, scale-to-zero, and AWS credits make it easier to test production-style search without ops overhead.
- Provisioning happens inside Vercel; env vars are injected automatically, cutting setup friction for search-backed apps.
- One collection supports vector, lexical, hybrid, and agentic search, so RAG and normal site search can share infra.
- Claims include 20x faster autoscaling, scale-to-zero, and up to 60% cost savings versus peak-capacity pricing.
- New AWS accounts created through Vercel get $100 credits; existing AWS accounts can be linked and monitored from the dashboard.
Source: vercel.com/changelog/amazon-opensearch-serverless-is-nowRead original →
FIG-8251:1
70radar
FIG-8251:1
#0824
#0824Infra & SaaS vercel_blog3 days ago
`Vercel AI Gateway` Adds Team-Wide Provider Allowlist
60radar
AI GatewayAI routing SaaS — one API across multiple model providers
Provider routing can now be locked at the account layer. Useful when client, compliance, or cost rules require approved AI vendors only.
- The allowlist applies to every AI Gateway request, including BYOK traffic. Request-level filters cannot bypass it.
- Filtering is by provider, not model. If one provider fails, Gateway can fall back only to another allowed provider for that model.
- New providers are disabled by default once the allowlist is on, so vendor scope will not expand silently.
- Only team owners can change the list. Coding agents are also blocked from routing to unapproved providers.
Source: vercel.com/changelog/team-wide-provider-allowlist-on-ai-Read original →
FIG-8241:1
60radar
FIG-8241:1
#0823
#0823Infra & SaaS GeekNews3 days ago
`Cloudflare Flagship` Launches Feature Flags for Workers
70radar
Cloudflare FlagshipFeature flag service — native evaluation inside Workers
Feature exposure can now change without redeploys. Native Workers evaluation makes rollout control cheap to wire in for edge apps.
- Flags are defined with targeting rules and percentage rollouts, covering staged releases and cohort-based experiments without app redeploys.
- Native Workers bindings evaluate flags directly, so edge-side feature gating avoids an extra feature-flag API hop.
- OpenFeature support matters: less vendor lock-in if the project already uses a standard flag interface.
Source: news.hada.io/topic?id=29934Read original →
FIG-8231:1
70radar
FIG-8231:1

Wed, May 273 dispatches

#0822
#0822Infra & SaaS GeekNews4 days ago
`Ghost CMS` Sites Hit by `ClickFix` Campaign via Critical CVE
60radar
Ghost CMSOpen-source CMS — supports newsletters and memberships
A critical CMS flaw is being used to turn legitimate sites into fake security-check traps. If you run Ghost, patch and audit immediately.
- Attackers abused CVE-2026-26980, described as a critical Ghost CMS vulnerability; public disclosure quickly became live exploitation.
- More than 700 sites were reportedly infected, so this has moved beyond theoretical risk into active campaign territory.
- ClickFix tricks visitors with fake security verification flows; compromised content can damage trust even if your app backend stays intact.
Source: news.hada.io/topic?id=29916Read original →
FIG-8221:1
60radar
FIG-8221:1
#0821
#0821Infra & SaaS stripe_blog4 days ago
`Stripe Radar` expands fraud protection beyond card payments
80radar
Stripe RadarPayment fraud detection SaaS — risk blocking on Stripe's network
Fraud screening now covers all supported payment methods, multi-account abuse, and pay-as-you-go abuse. Useful if billing risk spans subscriptions, credits, or marketplace sellers.
- High-risk transactions can now be blocked across all supported payment methods, so non-card checkout paths get the same fraud layer.
- Radar now targets multi-account abuse and pay-as-you-go abuse, which matters for credit-based SaaS, trials, and usage billing.
- Protection can apply even when another payment processor is involved. That makes Radar closer to a standalone risk layer.
- Platforms get new merchant-risk tools for activity on and off Stripe, useful for marketplaces and creator SaaS with seller onboarding.
Source: stripe.com/blog/expanding-stripe-radar-to-protect-more-oRead original →
FIG-8211:1
80radar
FIG-8211:1
#0820
#0820Infra & SaaS vercel_blog4 days ago
`Vercel CLI` Adds Experimental Native Binaries Without Node.js
60radar
Vercel CLIDeployment CLI — manages and deploys Vercel projects
Deploy commands can now run from signed native binaries with no Node runtime. Faster startup and OS-scoped credential storage make it worth testing on daily deploy machines.
- Install via @vercel/vc-native; -f is required because it uses the same global vercel and vc bin names as the standard CLI.
- Binaries are code-signed, so the OS can verify they came from Vercel and were not modified.
- On macOS, credentials move into the system Keychain scoped to the binary, reducing token exposure to other local processes.
- Support covers macOS, Linux, and Windows on x64 and arm64; useful for lean CI images and fresh machines.
Source: vercel.com/changelog/experimental-native-binaries-for-veRead original →
FIG-8201:1
60radar
FIG-8201:1

Tue, May 263 dispatches

#0819
#0819Infra & SaaS GitHub Changelog4 days ago
`GitHub` Adds PR Code Coverage Preview for Code Quality Users
70radar
Coverage now sits directly in the PR review loop. It reduces CI tab-hopping and makes test gaps harder to miss before merge.
- Available in public preview for all GitHub Code Quality users on github.com; coverage becomes part of normal PR review.
- PRs can show an aggregate covered-code percentage, so reviewers get a quick test-risk signal without opening separate reports.
- Best fit is repos already paying for GitHub Code Quality; teams outside that product get no immediate workflow change.
Source: github.blog/changelog/2026-05-26-code-coverage-in-pull-rRead original →
FIG-8191:1
70radar
FIG-8191:1
#0818
#0818Infra & SaaS vercel_blog5 days ago
`Vercel Sandboxes` Persistence Is Now GA
80radar
Vercel SandboxesCloud execution sandbox — automatic filesystem restore
Filesystem state now survives sessions by default. Long-running agent and preview workflows get simpler, but snapshot storage becomes a new cost line.
- Sandbox.create() now enables persistence by default, so files are saved and restored without manual snapshot handling.
- Each sandbox gets a customizable durable name. Sandbox.getOrCreate() adds idempotent retrieve-or-create for long-lived environments.
- Automatic snapshots consume separately billed snapshot storage. Use --non-persistent for throwaway jobs to avoid idle storage cost.
- Calling runCommand() or writeFiles() on a stopped sandbox resumes from the latest snapshot, which fits agent task handoff and preview reuse.
Source: vercel.com/changelog/sandbox-persistence-is-now-gaRead original →
FIG-8181:1
80radar
FIG-8181:1
#0817
#0817Infra & SaaS vercel_blog5 days ago
`Firecrawl` joins the `Vercel Marketplace`
60radar
FirecrawlWeb crawling API — extracts LLM-ready structured data
Teams can add crawling infrastructure directly from the hosting stack. It turns web pages into LLM-ready data, useful for RAG and agent features without running scrapers yourself.
- Firecrawl can scrape pages into markdown, HTML, structured data, or screenshots, cutting the glue code around ingestion.
- Search plus full-page retrieval works in one call, a practical fit for small RAG tools and monitoring agents.
- Dynamic page interaction via AI prompts or code reduces the pain of JavaScript-heavy sites, but usage cost still needs checking.
Source: vercel.com/changelog/firecrawl-joins-the-vercel-marketplRead original →
FIG-8171:1
60radar
FIG-8171:1

Sat, May 232 dispatches

#0816
#0816Infra & SaaS GitHub Changeloglast week
Staged publishing and new install-time controls for npm
70radar
npm enhances supply-chain security with staged publishing and new --allow-* install flags. These features provide granular control over package sources, mitigating dependency attacks.
- staged publishing: Publish with npm publish --stage <name> and later promote with npm stage promote. This prevents accidentally tagging a pre-release as `latest`.
- --allow-* flags: New flags like --allow-file and --allow-remote let you explicitly whitelist installation sources during npm install, expanding on the existing --allow-git.
- Both features focus on hardening defenses against supply-chain attacks. It's a good practice to configure them in .npmrc even for solo projects with many dependencies.
Source: github.blog/changelog/2026-05-22-staged-publishing-and-nRead original →
FIG-8161:1
70radar
FIG-8161:1
#0815
#0815Infra & SaaS RevenueCat Bloglast week
RevenueCat launches `AI Toolkit` for subscription setup
60radar
RevenueCat AI ToolkitSubscription setup toolkit — starts setup inside coding assistants
Subscription setup can now start inside an AI coding assistant. Useful when billing plumbing is the bottleneck, but scope is still narrow from the available details.
- The stated job is simple: set up subscriptions from an AI coding assistant, reducing context switching around billing setup.
- RevenueCat stays the billing backend; the new surface is the assistant workflow, not a separate payment provider.
- Available detail is thin: no pricing, supported assistants, or platform list. Treat as worth checking, not a migration trigger.
Source: www.revenuecat.com/blog/company/ai-toolkit/Read original →
FIG-8151:1
60radar
FIG-8151:1

Thu, May 215 dispatches

#0814
#0814Infra & SaaS GitHub Changeloglast week
`GitHub` Issue Fields Enter Public Preview for All Organizations
60radar
Typed metadata is now native on issues. Small teams can standardize triage without extra trackers.
- Priority, Effort, and similar typed fields can now live directly on GitHub issues, reducing custom labels and project-board drift.
- Public preview covers all GitHub organizations on github.com, so teams can test it without waiting for an enterprise rollout.
- GitHub Enterprise Cloud with data residency is included too; compliance-bound org repos get the same workflow option.
Source: github.blog/changelog/2026-05-21-issue-fields-are-now-inRead original →
FIG-8141:1
60radar
FIG-8141:1
#0813
#0813Infra & SaaS vercel_bloglast week
`Vercel CLI` Can Now Pull Anomaly Alert Details
60radar
Anomaly status now reaches the terminal, including AI investigation output with --ai. Useful for incident triage, but gated behind Observability Plus.
- vercel alerts lists team or project alerts with start time, alert type, and active status; terminal triage gets less context-switching.
- --ai adds AI investigation results next to each alert, giving coding agents enough context to act from the CLI.
- Availability is limited to Observability Plus, so the value depends on whether observability spend is already justified.
Source: vercel.com/changelog/pull-anomaly-alert-details-using-thRead original →
FIG-8131:1
60radar
FIG-8131:1
#0812
#0812Infra & SaaS vercel_bloglast week
`Qwen 3.7 Max` lands on `Vercel AI Gateway`
70radar
Qwen 3.7 MaxLLM — built for coding and long-horizon agents
Alibaba’s agent-focused model is now callable through the AI SDK. Routing, retries, cost tracking, and BYOK make it worth a quick prototype test.
- Use model ID alibaba/qwen-3.7-max; existing AI SDK paths can compare it without provider-specific wiring.
- Positioned for coding, office workflow automation, and long-horizon autonomous execution. Add it to multi-file engineering bake-offs.
- Gateway features include usage and cost tracking, retries, failover, and intelligent routing. One API reduces provider-switching overhead.
- Built-in BYOK, observability, and custom reporting keep model experiments easier to measure across small products.
Source: vercel.com/changelog/qwen-3-7-max-now-available-on-verceRead original →
FIG-8121:1
70radar
FIG-8121:1
#0811
#0811Infra & SaaS vercel_bloglast week
Configure weighted traffic splits for Vercel Flags from the Vercel CLI
60radar
The new vercel flags split command allows for weighted traffic distribution for feature flags. This simplifies setting up simple A/B tests without leaving your terminal, useful for quick experiments.
- You can run the command interactively or pass flags for automated environments.
- This removes the need to switch to the Vercel dashboard for managing traffic splits, keeping you in the development flow.
- It's designed for splitting traffic between two variants, making it ideal for gradual rollouts or simple UI changes.
Source: vercel.com/changelog/configure-weighted-traffic-splits-fRead original →
FIG-8111:1
60radar
FIG-8111:1
#0810
#0810Infra & SaaS Latent Spacelast week
`Railway` Pushes an Agent-Native Cloud Narrative
70radar
RailwayCloud PaaS — Git-based app deploys and infra management
Own-metal infra and heavy coding-agent spend point to a tighter cloud/dev loop. Worth watching if deploy workflows move from PRs to agent-driven changes.
- Railway claims 3M users and 100K signups per week; distribution now matters as much as hosting features.
- Own-metal data centers signal margin control, not just Heroku-style UX. That can shape pricing and latency later.
- $200K+ coding agent spend is a strong signal that internal engineering workflows are being rebuilt around agents.
- “Death of PRs” frames deployment as agent-native operations. Small teams should watch whether review, rollback, and audit trails keep up.
Source: www.latent.space/p/railwayRead original →
FIG-8101:1
70radar
FIG-8101:1

Wed, May 206 dispatches

#0809
#0809Infra & SaaS RevenueCat Blog2 weeks ago
`Google Play Billing Library` `9.0`, billing and console updates from I/O 2026
80radar
Google Play Billing LibraryAndroid billing library — Google Play in-app purchase integration
Billing, analytics, protection, and Play Console all moved at once. Android subscription code should be reviewed now, not after a build rejection.
- Play Billing Library 9.0.0 is a major version, so billing code deserves a compatibility pass before the next Android release.
- The update bundle includes AI, Play Console, analytics, and protection changes. Store operations are expanding beyond payment integration.
- RevenueCat coverage matters because it usually translates Google Play billing changes into SDK and migration impact.
- Apps with Android subscriptions should watch RevenueCat and Google timelines together. Late migration turns into release risk.
Source: www.revenuecat.com/blog/engineering/play-billing-v9/Read original →
FIG-8091:1
80radar
FIG-8091:1
#0808
#0808Infra & SaaS vercel_blog2 weeks ago
`Vercel Chat SDK` Adds `callbackUrl` for Card Buttons and Modals
60radar
Vercel Chat SDKChat UI SDK — integrates Vercel workflows and adapters
Agent runs can now pause and resume from UI actions. Useful for approval flows in Slack or Teams without extra glue code.
- callbackUrl sends button-click or form-submit payloads to your endpoint, so workflows can resume from external UI events.
- Cards need a workflow webhook URL wired into each button’s callbackUrl; modals include submitted form data in the payload.
- Button callbacks work on most official-adapter platforms, while modal callbacks are limited to Slack and Teams for now.
Source: vercel.com/changelog/chat-sdk-now-supports-callback-urlsRead original →
FIG-8081:1
60radar
FIG-8081:1
#0807
#0807Infra & SaaS vercel_blog2 weeks ago
`Chat SDK` adds message subjects and direct adapter SDK access
60radar
Chat SDKBot SDK — adapters for GitHub, Linear, and Slack
Bots can now resolve the parent GitHub issue, PR, or Linear item from a mention. This makes workflow bots less glue-heavy and worth adopting for support or triage automation.
- message.subject returns title, status, URL, and typed parent payload for GitHub and Linear comments; Slack returns null.
- Subject lookup is cached per message, so repeated reads avoid extra provider API calls in the same handling path.
- Adapters now expose underlying platform SDKs directly; custom provider API calls no longer need separate client wiring.
Source: vercel.com/changelog/chat-sdk-adds-message-subjects-and-Read original →
FIG-8071:1
60radar
FIG-8071:1
#0806
#0806Infra & SaaS vercel_blog2 weeks ago
`Chat SDK` adds built-in `AI SDK` tools via `chat/ai`
70radar
Chat SDKVercel chat app SDK — built-in AI tool wiring
Agent wiring is now one call: createChatTools(chat). Approval-gated writes and preset-scoped tools reduce custom glue code for Vercel chat apps.
- createChatTools(chat) connects Chat SDK read/write actions into an agent, cutting boilerplate around chat state operations.
- Write tools default to requireApproval, so destructive agent actions start behind an approval gate.
- reader, messenger, and moderator presets construct only allowed tools, keeping tool exposure narrower by default.
Source: vercel.com/changelog/chat-sdk-now-includes-ai-sdk-toolsRead original →
FIG-8061:1
70radar
FIG-8061:1
#0805
#0805Infra & SaaS vercel_blog2 weeks ago
`Vercel AI Gateway plugin` connects WordPress to 40+ AI providers
80radar
Vercel AI Gateway pluginWordPress AI connector — one-key routing to 40+ providers
WordPress sites can now route AI features through one gateway key instead of per-provider setup. Useful for shipping AI plugins with less credential, fallback, and billing plumbing.
- The connector works with the new WordPress AI Client, which requires WordPress 7.0; older sites need an upgrade first.
- One AI Gateway key in Settings > Connectors replaces separate credentials for Anthropic, Google, OpenAI, xAI, DeepSeek, and more.
- Plugins built on WordPress AI Client inherit the connector automatically, reducing custom provider integration work.
- Supported modes include text, structured JSON, image generation/editing, and video through the same prompt builder.
- Automatic fallbacks and dynamic model discovery lower outage handling and model-update maintenance.
Source: vercel.com/changelog/vercel-ai-gateway-plugin-for-wordprRead original →
FIG-8051:1
80radar
FIG-8051:1
#0804
#0804Infra & SaaS Hacker News · MCP Server2 weeks ago
`Superlog`, self-installing observability with bug-fix PRs
60radar
SuperlogObservability SaaS — auto-instruments and opens fix PRs
Observability setup moves from dashboard wiring to repo-scanning instrumentation. The promise is strong for small teams, but pricing and real-world accuracy decide whether it replaces Sentry or Datadog.
- The wizard scans a repo and adds structured logs, traces, and metrics through OpenTelemetry; setup pain is the main wedge.
- Incidents are fingerprinted and grouped, so alert flow becomes one issue instead of duplicate spam in Slack.
- The agent investigates errors and opens a tested PR when context is enough; otherwise it posts findings and pulls in likely owners.
- Cost tracking includes LLM/upstream spend by callsite, tenant, and model, useful for multi-tenant SaaS margins.
Source: superlog.sh/Read original →
60radar
PHOTO
FIG-8041:1

Tue, May 194 dispatches

#0803
#0803Infra & SaaS vercel_blog2 weeks ago
`Vercel` opens `Flat Rate CDN` limited beta for Pro teams
100radar
Flat Rate CDNVercel CDN plan — fixed monthly billing regardless of traffic
CDN bills can now be capped with a fixed monthly fee. This reduces launch-risk from viral traffic, bot spikes, and route mistakes.
- Flat Rate CDN replaces usage-based CDN billing with a fixed monthly price for Pro teams in limited beta.
- Edge Requests, Fast Data Transfer, and related Vercel CDN products are no longer metered under the plan.
- Temporary traffic spikes will not raise the bill, which makes launches and experiments easier to budget before the month starts.
Source: vercel.com/changelog/flat-rate-cdn-in-limited-betaRead original →
FIG-8031:1
100radar
FIG-8031:1
#0802
#0802Infra & SaaS vercel_blog2 weeks ago
`Vercel Sandbox` Now Runs `Claude Managed Agents`
80radar
Vercel SandboxIsolated runtime — per-session Firecracker microVMs
Agent tool calls can run inside isolated Vercel microVMs while secrets stay outside the sandbox. Useful for production agent features that need private APIs and tighter network control.
- Claude Managed Agents provides the model, harness, tools, and session state; Vercel supplies the execution environment for self-hosted tool calls.
- Each session gets its own Firecracker microVM with millisecond startup, making per-user agent isolation practical.
- Credential brokering happens at the firewall, so secrets do not enter the sandbox. This is the right default for customer-data workflows.
- Deny-by-default egress with a domain allowlist reduces blast radius when agents call tools or browse controlled services.
Source: vercel.com/changelog/run-claude-managed-agents-with-vercRead original →
FIG-8021:1
80radar
FIG-8021:1
#0801
#0801Infra & SaaS vercel_blog2 weeks ago
`Vercel` stops charging for WAF-mitigated traffic
90radar
VercelFrontend cloud platform — integrated Next.js deploys and CDN
Blocked bots no longer turn into surprise CDN bills. This directly lowers the risk of scrapers, credential stuffing, and abused API endpoints on small apps.
- Traffic denied, challenged, or rate-limited by Vercel Firewall now waives CDN Requests and Fast Data Transfer charges.
- DDoS mitigation was already unlimited and free; this change covers non-DDoS abuse handled by custom rules, managed rules, and rate limits.
- The waiver applies automatically to every project using Vercel Firewall; no config change or migration required.
- Product pages, login routes, and expensive API endpoints are the obvious places to add stricter rules now.
Source: vercel.com/changelog/web-application-firewall-mitigated-Read original →
FIG-8011:1
90radar
FIG-8011:1
#0800
#0800Infra & SaaS RevenueCat Blog2 weeks ago
`RevenueCat` Adds Flexible Discounts for Web Billing
60radar
RevenueCatSubscription billing infra — unified app and web subscriptions
Web subscriptions can now use percentage discounts, promo codes, and win-back offers. Useful for churn recovery and price testing without wiring discount logic yourself.
- Percentage-off discounts and promo codes now apply to RevenueCat web subscribers, reducing custom billing work.
- Win-back offers target cancelled or lapsed users. That gives churn recovery a built-in path instead of ad hoc email-only campaigns.
- The scope is web billing, not App Store pricing. Best fit is a paid web checkout where discount tests can move quickly.
Source: www.revenuecat.com/blog/company/flexible-discounts-web-bRead original →
FIG-8001:1
60radar
FIG-8001:1

↻ Load earlier issues

`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

`Vercel Sandbox` now runs Docker containers

`Amazon OpenSearch Serverless` lands in the `Vercel Marketplace`

`Vercel AI Gateway` Adds Team-Wide Provider Allowlist

`Cloudflare Flagship` Launches Feature Flags for Workers

`Ghost CMS` Sites Hit by `ClickFix` Campaign via Critical CVE

`Stripe Radar` expands fraud protection beyond card payments

`Vercel CLI` Adds Experimental Native Binaries Without Node.js

`GitHub` Adds PR Code Coverage Preview for Code Quality Users

`Vercel Sandboxes` Persistence Is Now GA

`Firecrawl` joins the `Vercel Marketplace`

Staged publishing and new install-time controls for npm

RevenueCat launches `AI Toolkit` for subscription setup

`GitHub` Issue Fields Enter Public Preview for All Organizations

`Vercel CLI` Can Now Pull Anomaly Alert Details

`Qwen 3.7 Max` lands on `Vercel AI Gateway`

Configure weighted traffic splits for Vercel Flags from the Vercel CLI

`Railway` Pushes an Agent-Native Cloud Narrative

`Google Play Billing Library` `9.0`, billing and console updates from I/O 2026

`Vercel Chat SDK` Adds `callbackUrl` for Card Buttons and Modals

`Chat SDK` adds message subjects and direct adapter SDK access

`Chat SDK` adds built-in `AI SDK` tools via `chat/ai`

`Vercel AI Gateway plugin` connects WordPress to 40+ AI providers

`Superlog`, self-installing observability with bug-fix PRs

`Vercel` opens `Flat Rate CDN` limited beta for Pro teams

`Vercel Sandbox` Now Runs `Claude Managed Agents`

`Vercel` stops charging for WAF-mitigated traffic

`RevenueCat` Adds Flexible Discounts for Web Billing