Weekly digest
What shipped, week by week
The digest is the newsletter — same content, sent every Sunday. This page is the public archive: every weekly send, browsable, no signup required to read.
Get it in your inbox
Weekly + when something material ships. No daily spam, no third-party article aggregation.
- 4 updates2026-W38
September 14–20
- contentSep 18Deals of the Week refreshed: 10 hand-picked AppSumo lifetime deals
- contentSep 17New post: measure what your agent harness sends on every call
- contentSep 16Three tools added: Hermes Agent, Context7, Headroom
- contentSep 16New post: two CISA deadlines in the Python packages under your LLM proxy
- 7 updates2026-W37
September 7–13
- contentSep 11Deals of the Week refreshed — 9 hand-picked AppSumo deals
- contentSep 11New post: deepseek-v4-pro answers with a different model on Monday
- contentSep 10New post: the "use TDD" line in your AGENTS.md scored below writing nothing
- contentSep 9Three new tools: SkillOpt, Graft, OpenConnector
- contentSep 8Two skills: holding an agent to the smallest solution that works, and diagrams built to be looked at
- contentSep 7Three prompts: SLOs and an error budget policy, an A/B test pre-registration, and a design token audit
- contentSep 7A post on auditing what your agent skills cost, and pinning where they came from
- 10 updates2026-W36
Aug 31 – Sep 6
- fixSep 6Every skill pack now says what it holds, in the words people search
- contentSep 6New post: the sampling parameters left the frontier APIs, and effort is not a drop-in
- contentSep 4Deals of the Week refreshed — 10 hand-picked AppSumo deals
- contentSep 4New post: the 403 that GPT-6 Astra can return mid-run, and why your retry layer must not touch it
- contentSep 3New post: four checks that find what a release contains
- contentSep 2Three tools: an agent that keeps its own memory, AI coworkers that ask before acting, and a wiki that rewrites itself
- contentSep 2New post: what breaks when you point existing code at Claude Fable 5.1
- contentSep 1Two skills: a written quality bar agents cannot quietly lower, and a scroll page that verifies its own scroll
- contentAug 31Three prompts: a STRIDE threat model, an OWASP review of an AI feature, and a data contract
- contentsecurityAug 31New post: logging out everywhere does not reach the credential your coding agent stored
- 7 updates2026-W35
August 24–30
- contentAug 30New guide: prompting an AI app builder for a real backend
- contentAug 30New post: seven dated changes landed this week, and one is on a deprecation page
- contentAug 29New post: OpenAI models leave Cursor on November 12
- contentAug 28Deals of the Week refreshed — 10 hand-picked AppSumo deals
- contentAug 27New post: poll the model provider’s status page, not your coding tool’s
- contentAug 24Three prompts: alert triage, live incident updates, and a ranked cloud cost list
- contentAug 24New post: crop the screenshot before you send it
- 9 updates2026-W34
August 17–23
- fixcontentAug 23Search housekeeping: HR skill pack title, role-prompting guide, 23 internal links repointed
- contentsecurityAug 23New post: what an agent sandbox rule does not cover
- contentAug 21Deals of the Week refreshed — 10 hand-picked AppSumo deals
- contentAug 20New post: query your real cost per million tokens instead of reading the price list
- contentAug 19Three agent-runtime tools added to the catalogue
- contentAug 18Two new skills: naming the tells in AI-generated UI, and vetting a skill before you install it
- contentAug 18New post: Cursor now hosts your code, and the retention terms are not published yet
- contentAug 17Three prompts: an eval grader, month-end variance commentary, and one lesson for a mixed class
- contentAug 17New post: Imagen 4 is gone, and the obvious replacement is the expensive one
- 12 updates2026-W33
August 10–16
- fixAug 16Three search snippets rewritten to match what people actually type
- contentAug 16New post: a price per million tokens is no longer one number
- contentAug 14Deals of the Week refreshed — 10 hand-picked AppSumo deals
- contentAug 14New post: DeepSeek's off-peak rate is not a discount
- contentAug 13New post: two Claude Code hooks that catch a broken build before you do
- contentAug 13New post: Qwen3.8-Max's open weights landed, license attached
- contentAug 12Three new tools: Goose, Microsoft Agent Framework, Rivet
- contentAug 12New post: Claude output is watermarked now, and what a detected mark actually proves
- contentAug 11Two new skills: draw.io diagrams from real artifacts, docs in Simplified Technical English
- contentAug 11New post: the four checks before moving an agent loop onto Muse Glimmer
- contentAug 10Three new prompts: SEO content briefs, user interviews, jobs-to-be-done
- contentAug 10New post: cut the tokens Claude Code spends before it reads your prompt
- 10 updates2026-W32
August 3–9
- fixAug 9Search snippets corrected on three pages
- contentAug 9New post: frontier labs now treat their own models as insider threats
- contentAug 7Deals of the Week refreshed: 12 hand-picked AppSumo deals
- contentAug 7New post: Agent Plugins 1.0, one package for six agent clients
- contentAug 6New post: turn a PDF into an agent skill it loads on demand
- contentAug 6New post: Muse Code and the tier that trains on your code
- contentAug 5Three new tools: Buzz, OpenWork, Screenpipe
- contentAug 4Two new skills: marketing attribution and performance optimization
- contentAug 3Three new prompts: Excel formulas, AI usage policy, 90-day onboarding
- contentAug 3New post: MCP went stateless — the server migration checklist
- 8 updates2026-W31
Jul 27 – Aug 2
- contentAug 2New post: the EU AI Act deadline that was not postponed
- contentAug 2New post: two model IDs change meaning on August 5
- contentJul 31Deals of the Week refreshed: Prompt Architects added
- contentJul 30New post: the fallback pattern for when Claude or OpenAI's API goes down
- contentJul 29Three tools added: Chrome DevTools MCP, Orca, Graphify
- contentJul 28New skill: checking marketing copy for AI-sounding tells
- contentJul 273 new prompts: contract risk review, RACI matrix, support burnout signals
- contentJul 27New post: why a Claude Code skill stops triggering, and the description fix
- 7 updates2026-W30
July 20–26
- contentJul 26New post: what a ransomware attack and a sandbox escape both prove about agent security
- contentJul 24Deals of the Week refreshed — 12 hand-picked AppSumo deals
- contentJul 23New post: activating Claude Security's multi-agent vulnerability scan
- contentJul 22Three tools added: opencode, Browser Use, Unsloth
- contentJul 22New post: what OpenAI's sandbox-escape postmortem means for your own agents
- contentJul 21New skill: planning a free tool as a marketing channel
- contentJul 21New post: how to actually use Kimi K3 today, and what to hold off on
- 5 updates2026-W29
July 13–19
- contentJul 17Deals of the Week refreshed — 11 hand-picked AppSumo deals
- contentJul 16New post: your prompt cache can miss on every request without warning
- contentJul 142 new skills: API contract design and app store listing audits
- contentJul 133 new prompts: peer review response, query letters, self-editing
- contentJul 13New post: catch your AI's own mistakes with an independent verifier subagent
- 9 updates2026-W28
July 6–12
- contentJul 12New post: the $9 billion forward-deployed-engineer race
- contentJul 10Deals of the Week refreshed
- contentJul 10New post: ChatGPT Sites public beta, what it replaces
- contentJul 9New post: route Claude Code subagents to cheaper models
- contentJul 9GPT-5.6 GA day: the Terra migration test to run
- contentJul 83 new tools: Dify, Firecrawl, ComfyUI
- contentJul 72 new skills: Implementation Plan Writer, Figma Design Implementer
- contentJul 63 new prompts: mutual action plan, deal qualification gap-check, retirement plan stress-test
- contentJul 6New post: search your coding agent's own session history instead of re-explaining it
- 8 updates2026-W27
Jun 29 – Jul 5
- fixJul 5emergentintegrations guide title now names the Python package
- contentJul 5New post: frontier AI is going metered — budget by task, not by seat
- contentJul 3Deals of the Week refreshed — 10 hand-picked AppSumo deals
- contentJul 2New post: automate a competitor watch with a Claude Code routine
- contentJul 12 new tools: Skybridge, Hyper
- contentJul 1New post: Claude Sonnet 5 cost and when to switch
- contentJun 293 new prompts: salary negotiation, character arc, workflow automation
- contentJun 29New post: How to get reliable JSON from Claude and GPT
- 13 updates2026-W26
June 22–28
- contentJun 28New guide: emergentintegrations.llm.chat — how it works and JavaScript alternatives
- contentJun 28New post: GPT-5.6 Sol, Terra, and Luna — confirmed pricing and what the staged rollout means
- contentJun 26New guide: How to create an AI agent
- contentJun 26Deals of the Week — 7 hand-picked AppSumo lifetime deals
- contentJun 25New: AI app builder comparisons (Lovable, Bolt, v0, Replit, Emergent, Base44)
- contentJun 25Spotlight: Lovable — full-stack apps from a prompt
- contentJun 25New post: OpenAI's first custom chip — what Jalapeño means for inference costs
- contentJun 243 new design prompts: Creative Brief, Heuristic Evaluation, Micro-Interaction Spec
- contentJun 242 new skills: Doubt-Driven Development + Browser Testing with DevTools
- contentJun 243 new tools: Wispr Flow, Steel, Baseten
- contentJun 22Deal: WriteVoice lifetime on AppSumo
- contentJun 222 new skills: Excalidraw Diagram Builder + OKR Design Facilitator
- contentJun 223 new tools: Viktor, Octolane, MiniMax Hub
- 17 updates2026-W25
June 15–21
- contentJun 21New post: Build a 20-case eval set for your AI feature (with promptfoo)
- contentJun 212 new skills: Opportunity Solution Tree (founder) + Jobs to Be Done Analyzer (marketer)
- contentJun 213 new prompts: architecture risk review, Feynman learning loop, voice-preserving editor
- contentJun 213 new tools: Poolside, API to MCP, Atomic Mail Agentic
- contentJun 21New post: Claude Fable 5 suspended — what to do if you build on Claude
- featureJun 21Every prompt now has its own page
- contentJun 21Learn: two more migration guides — Bolt and v0
- contentJun 21Learn: two new pillars + a decision hub — choosing AI, and running the tools
- contentJun 21Learn: the "Building with prompts" pillar is now live (4 guides)
- contentJun 202 new skills: Threat Model Builder (developer) + Investor Outreach Sequencer (founder)
- contentJun 203 new prompts: sprint retro, blameless post-mortem, SWOT with priorities
- contentJun 20165 new tools across all 12 categories
- contentJun 18New post: Gemini CLI stops serving Pro, Ultra, and free users today
- contentJun 16New post: SpaceX is acquiring Cursor for $60 billion
- contentJun 167 new prompts: research, newsletter, launch, investor updates
- contentJun 16Best AI for your role — 9 more pick lists
- contentJun 16Best AI for your role — 7 new pick lists
- 4 updates2026-W24
June 8–14
- 4 updates2026-W23
June 1–7
- 19 updates2026-W22
May 25–31
- contentMay 31New post: How to write AGENTS.md so AI doesn't break your repo
- contentMay 31New post — How to write AGENTS.md so AI doesn't break your repo
- contentMay 31Guide expanded — Base44 migration FAQ
- contentMay 27New pillar content — four AI-builder migration guides
- contentMay 26+65 new prompts and +43 new skills added to the library
- contentMay 26End-of-month refresh — 3 new spotlights, /best surfaces reviewed
- contentMay 268 new tools in /tools + new guide on evaluating indie AI tools
- featurecontentMay 26Two more learning paths + in-path indicator on guide pages
- contentfeatureMay 26New foundations pillar + learning paths feature in /learn
- contentMay 26Curated guide — free AI courses worth your time
- contentMay 26New /learn pillar — Prompt craft — with 3 foundational guides
- contentMay 2618 new skills in /skills — founder, sales-rep, developer, marketer
- contentMay 2626 new prompts across /sales, /productivity, /writing, /research, /design, /analysis
- contentMay 2614 new code prompts in /prompts — agent steering, incident response, planning, IaC review
- contentMay 25Phase 1B: four task-based /learn guides for using AI at work
- designfixMay 25SEO + brand cleanup: new favicon, fixed dark-mode contrast, edge-to-edge aurora, 830 fewer 404s/week
- contentMay 25RAG: sharper glossary entry + new in-depth /learn guide
- contentfeatureMay 25Learn restructured: 4 new cornerstone guides + new "Verify and trust" pillar
- contentMay 25New /learn guide: Best AI prompts for sales reps
- 2 updates2026-W21
May 18–24
- 2 updates2026-W20
May 11–17
- 8 updates2026-W19
May 4–10
- contentMay 107 new code prompts on verification + agent workflows
- contentMay 10New post: Before you ship — a pre-deploy checklist for AI-built apps
- featureMay 10SEO + AEO upgrade — llms.txt, JSON-LD everywhere, per-post OG images
- designMay 10Newsletter template redesigned
- contentMay 7New post: Most AI-built apps shouldn't be on the internet
- featurecontentMay 5/open hub + OpenClaw + Posts rename
- contentfeatureMay 54 new skills + 5 new prompts on working safely with AI
- contentMay 4New post: Your AI is confidently wrong
- 3 updates2026-W18
Apr 27 – May 3
Don't want to come back here?
Get the digest in your inbox every Sunday.