Changelog

Added

• The App shell code pack: a runnable Next.js app frame with working sign-in, a protected members area, page titles and sitemap handled, and a live setup report. Install it into an empty folder with `npx -y boostmyagent starter app`.
• It runs the moment it installs, with zero vendor accounts. Built-in stand-ins cover sign-in, storage, email, and analytics while you build, so nothing blocks your first day.
• Every vendor decision stays open at a named plug-in point (a seam). When you are ready, wire sign-in from the Supabase, Clerk, or Firebase pack, payments from Stripe or Whop, email from Resend or Mailgun, analytics from GA4 or PostHog, errors from Sentry, and hosting from Vercel or Cloudflare.
• A live setup report at /api/verify tells you in plain language what is wired, what still runs on a stand-in, and which file to open next.
• Same release bar as every pack: six different AI coding models adapted it without help, across twelve separate test runs, before it shipped.

Changed

• The code directory at `/code` and the homepage now lead with the App shell, so new builders see the starting point before the packs that plug into it.

Added

• 17 new code packs, all installable today with one command: Supabase, Firebase, Clerk, Cloudflare, Vercel, OpenAI, Claude, Gemini, MCP, GA4, Resend, Mailgun, Sentry, PostHog, Expo, RevenueCat, and Whop.
• Every pack passed the same bar before release: six different AI coding models adapt it without help, and it ships only when they all succeed.
• Each pack is checked against its vendor's current documentation on the release date, so your agent builds with today's APIs instead of the ones it remembers from training.
• The /os menu now shows the code packs your license includes, with sizes and the install command for each.

Changed

• The code directory at `/code` now lists every module and file across all 18 packs, so you can browse everything before you buy.

Added

• Agentic commerce accepts agent-initiated payments through Shared Payment Tokens (Stripe credentials limited by seller, amount, currency, and expiration).
• Usage billing reports AI tokens or requests to Stripe meters so customers can pay for what they use.
• Customer portal sessions let customers manage plans, payment methods, and cancellations. Refund and dispute handling keeps access and payment records aligned.

Changed

• The Stripe pack now spans 12 modules, with every file path, purpose, language, and line count listed at `/code/stripe`.

Added

• The first code pack covers Stripe checkout, signed webhooks (messages Stripe proves it sent), subscriptions, fulfillment, emails, verification, shared helpers, and setup.
• Browse every included module, file path, purpose, language, and line count at `/code`. Source contents stay private to license holders.
• Install the Stripe pack inside your project with `npx -y boostmyagent starter stripe`. Firebase and Supabase packs will ship to every license holder when ready.

Added

• When a newer Boost OS version is available, your coding tool adds one line at the end of the session telling you to run /update.
• During the first seven days after install, /os opens with three starter steps: check your current work with /vibe, save what the session learned with /remember, and check your deployed service for errors with /health.
• /recall can now open synced memory (notes backed up to your Boost OS account) from another project when no local reference matches. It only reads and prints the reference into chat. Nothing is written to the current project.
• The memory view on /access lets you browse synced projects and open their saved reference files from your account.

Changed

• Synced memory now shows each project's readable name instead of an internal code.
• When a request belongs in a hosted service, external dashboard, or manual workflow, the agent gives short numbered steps for where to complete it and does not touch code.
• Starting with installer version 1.2.0, the usage counter is on by default for new installs. It counts command names only, never file contents or paths. Turn it off any time with `npx -y boostmyagent telemetry off` (telemetry is the technical name for this counter).

Added

• /pause turns off the parts of Boost OS that act on their own: the always-on guards (checks your coding tool runs automatically) and the instructions file read at the start of every chat. Your slash commands, project memory, and saved notes stay exactly where they are. Nothing is deleted, so there's no confirmation step.
• /resume reconnects the guards and puts the instructions file back exactly as it was. If new instructions appeared in your project while Boost OS was paused, it keeps both files and tells you which two to look at instead of overwriting either one.
• Boost OS now ships twenty-two slash commands.

Changed

• Updating a paused project keeps it paused. Installing a new release won't quietly turn the guards back on.
• /uninstall now suggests /pause first when you only want Boost OS quiet for a while, not gone.

Fixed

• The memory reminder (the note asking your agent to save what the session learned) was meant to appear once per chat, but it kept coming back after every reply while you had unsaved work. It now stays quiet for the rest of that chat, while new chats still get one reminder. Pick up the fix with `npx -y boostmyagent@latest update`.

Added

• A push guard checks every outgoing change for secret keys before your agent sends it to the shared project. It blocks the send if a secret slipped in before the commit guard (a check before saving code changes) was watching.
• An install guard warns when a library name is one typo away from a well-known library, a common trick for spreading malicious code, or when the library is less than 30 days old. It warns and steps aside, so the install still runs. Boost OS now ships six always-on guards.
• A settings guard checks your private settings file (`.env`) against its shared example copy and points out any setting that exists in one but not the other. This catches a quiet mistake that can break the next time your app goes live.
• An optional usage counter, off by default, counts which slash commands you run. It never counts what you typed, your files, or their contents. Turn it off any time.

Changed

• The repeatable checks inside /vibe, /security, and /sync now run in the Boost OS installer instead of relying on written instructions the AI follows. They still scan for secrets, leftover testing lines, unsafe settings files, and outdated libraries. If the installer is unavailable, the commands fall back to the old way.
• Multiple guards can now watch the same action without overwriting each other. Existing installs pick up all six guards on the next update.

Added

• /whats-new shows what shipped since your installed version, in plain bullets, with the one-line update command at the end. It only reports, so nothing installs until you say so. Boost OS now ships twenty slash commands.
• The /os menu now ends with a short "start here" suggestion matched to your project: brand-new projects get pointed at /design and /explain, active products at /vibe and /security.
• Installing Boost OS now marks the risky parts of your app as hands-off in project memory. These include payment webhooks (code that hears back from your payment provider), login checks, and database change history. The always-on guard warns your agent the moment it touches them.

Changed

• The install now finishes with a three-step first-five-minutes checklist (/os, then /vibe, then /remember) so you know exactly what to type first.
• The secret-key patterns that /vibe, /security, and the commit guard (a check before saving code changes) scan for now live in one shared file. A new leak pattern reaches all three at once. GitHub login tokens (secret codes that give account access) joined the list.
• The /sync command's catalog of outdated-code patterns now ships separately, so it can grow with every release.
• The installer is better at telling us which AI tool you use (anonymously, for prioritizing fixes) when you run it from a terminal (the window where you type commands).

Fixed

• The commit guard (a check before saving code changes) and the other two guards waited for a signal Cursor never sends before answering, which could freeze every command your agent tried to run. Guards now read the message as it arrives and answer immediately.
• Guard entries written to Cursor's settings now include a 10-second time limit, so even a misbehaving guard gets skipped instead of freezing anything.
• Already installed v1.1.0 in Cursor and seeing commands hang? Run `npx -y boostmyagent@latest update` in your computer's command window (not the agent's chat), then reload the Cursor window.

Added

• Always-on hooks (checks your coding tool runs automatically, no command needed). Three ship today: a commit guard that stops a commit (a saved snapshot of your code) when it contains secret keys, env files (the files holding your app's private settings), or leftover markers from two clashing edits; a fragile-file guard that warns the agent the moment it edits a file you've marked hands-off in project memory; and a memory nudge that reminds the agent to save what the session learned before it wraps up.
• Hooks connect automatically at install for the tools that support them: Cursor, Claude Code (GitHub Copilot reads the same setup), Codex, and Gemini CLI (Gemini's coding tool for terminal commands). Your existing hooks stay untouched. Skip with `--no-hooks` at install, or remove any time with `npx -y boostmyagent hooks uninstall`.
• Memory sync (backing up project notes to your account) uses `npx -y boostmyagent memory push` to save your notes, `memory pull` to restore them on another computer, and `memory status` to tell you whether the two copies match. Installing Boost OS on a new computer restores your latest backup automatically. Nothing on your computer gets overwritten without your say-so.

Changed

• /remember now also backs up your notes to your account after saving them on this computer. If you're offline or signed out, it skips the backup quietly. The save on your computer always wins.

Added

• /mobile is a Check command. Point it at one screen by file path, page address, or screen name, and it checks for hard-to-read text, tap targets too tiny for a thumb, sideways scrolling, images that spill off the edge, and content hidden behind the notch. You get a numbered to-do list for /preflight, and no code is changed.

Changed

• Boost OS now ships nineteen slash commands. Build: /cost, /design, /explain, /next, /spawn. Check: /audit, /health, /mobile, /preflight, /security, /sync, /vibe. Memory: /forget, /recall, /remember. Customize: /new, /uninstall, /update.

Added

• /design is a Build command. It walks you through four steps: define the feature in one sentence, see which files it touches, pick the simplest version, and create a numbered plan for /preflight before any code gets written.
• /security is a Check command. It looks for secrets in current work and the last 50 commits (saved snapshots of your code), checks that private settings files stay private, reports installed libraries with serious security warnings, and spots risky server addresses that other software can call. Reports hide the actual secret values.
• /sync is a Check command. It shows which installed libraries can be updated safely, which need review, which changed substantially, and which appear abandoned. For large version gaps, it reads the release notes and points out old library instructions still used in your code. The result is ready to paste into your next prompt.

Changed

• Boost OS now ships eighteen slash commands. Build: /cost, /design, /explain, /next, /spawn. Check: /audit, /health, /preflight, /security, /sync, /vibe. Memory: /forget, /recall, /remember. Customize: /new, /uninstall, /update.

Changed

• /update now runs `npx -y boostmyagent@latest update` after checking that Boost OS is installed, then prints what the installer actually said.
• /update no longer carries fake example versions or tries to interpret release notes in its own instructions. Release details come from the installer, not a guess.

Changed

• When the chat already has concrete findings, /remember keeps its one-note-at-a-time flow. When the chat starts without earlier findings, it inspects the project first and fills in the reference file's standard sections.
• A broad /remember request can split the work across up to five helpers working at the same time, after asking you first, so wide feature areas get mapped before the reference file is written.

Added

• /uninstall: asks before removing anything, then runs the installer's uninstall path from chat.
• Installer: `boostmyagent uninstall` removes the Boost OS command folders, the always-on rules file (`AGENTS.md`), project memory, references, the install record, and any empty folders left behind. Your saved license stays so you can install in other projects, unless you pass `--logout`. Custom commands you built with /new also stay.

Changed

• Boost OS now ships fifteen slash commands, with /uninstall in the Customize category beside /new and /update.

Added

• /next: new Build command. Reads the recent chat and drafts 2-3 copy-pasteable prompts you could send next, each written in plain language with technical terms explained inline. Built for the moment you know what you want but not how to phrase it.

Changed

• Check commands (/spawn, /health, /preflight, /vibe, /audit) now end with a short "Next moves" list tied to the actual findings, including specific files, commands, and decisions. A report now tells you what to do about it.

Added

• The kernel (the always-on rules file `AGENTS.md`) ships with every install. It tells the agent to read project memory first, stop on requests with no code to change, preserve your changes, debug from evidence, and route everything else through /os.
• Build category: /cost compares money, build time, learning curve, how hard it is to switch providers, and maintenance. /explain rewrites agent output in plain language, /next drafts your next prompt, and /spawn splits a task across helpers working at the same time.
• Check category: /vibe checks changes not yet saved in your project history for leaked secrets, broken builds, and risky design choices. /audit finds old and unused code with a yes/no prompt for each file, /preflight checks a plan against your existing code, and /health pulls recent production errors from services your app uses.
• Memory category: /remember, /recall, and /forget manage one reference note per feature area of your app.
• Customize category: /new builds custom slash commands through a short conversation, and /update installs newer commands, improvements to the always-on rules file, and bug fixes.
• The /os menu groups commands in workflow order: Build, Check, Memory, then Customize. Your custom commands appear in a separate "Your skills" section.
• Install creates project memory at `.agents/memory/project.md` from a one-time scan of your tools, where your app runs, commands that check the project, and feature areas.
• A blank reference file is created at install for every feature area the scan finds, ready to fill in.
• Installer commands: install, update, status, and logout. The installer also gives slash commands a focused way to read and write project memory.
• Install targets (the folder where commands are added): `--target auto` writes to `.agents/skills/` by default and falls back to `.claude/skills/` when the project only contains Claude setup files. `--target claude` and `--target agents` force the destination.
• Compatibility with Cursor, Claude Code, Codex, GitHub Copilot, Replit, Windsurf, Cline, and Aider.