Add Access Daily Journal From CLI With HTTP Server as a Logseq TIL

README.md

@@ -10,7 +10,7 @@ working across different projects via [VisualMode](https://www.visualmode.dev/).
 
 For a steady stream of TILs, [sign up for my newsletter](https://visualmode.kit.com/newsletter).
 
-_1884 TILs and counting..._
+_1895 TILs and counting..._
 
 See some of the other learning resources I work on:
 
@@ -59,6 +59,7 @@ If you've learned something here, support my efforts writing daily TILs by
 * [Kitty](#kitty)
 * [Linux](#linux)
 * [LLM](#llm)
+* [Logseq](#logseq)
 * [Mac](#mac)
 * [Math](#math)
 * [Mise](#mise)
@@ -169,9 +170,11 @@ If you've learned something here, support my efforts writing daily TILs by
 
 - [Allow Edits From The Start](claude-code/allow-edits-from-the-start.md)
 - [Ask An Agent To Fix Merge Conflicts](claude-code/ask-an-agent-to-fix-merge-conflicts.md)
+- [Build From My Phone In Cloud Environments](claude-code/build-from-my-phone-in-cloud-environments.md)
 - [Distinguish Sessions With Different Colors](claude-code/distinguish-sessions-with-different-colors.md)
 - [Monitor Usage Limits From CLI](claude-code/monitor-usage-limits-from-cli.md)
 - [Open Current Prompt In Default Editor](claude-code/open-current-prompt-in-default-editor.md)
+- [Pick Up Session Locally That Started In Cloud](claude-code/pick-up-session-locally-that-started-in-cloud.md)
 - [Resume Specific Session](claude-code/resume-specific-session.md)
 - [Set Permission Mode When Starting Session](claude-code/set-permission-mode-when-starting-session.md)
 - [Stash The Current Prompt To Send Another First](claude-code/stash-the-current-prompt-to-send-another-first.md)
@@ -227,6 +230,7 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Parameterized SCSS Mixins](css/parameterized-scss-mixins.md)
 - [Prevent Invisible Elements From Being Clicked](css/prevent-invisible-elements-from-being-clicked.md)
 - [:root Has Higher Specificity Than html](css/root-has-higher-specificity-than-html.md)
+- [Show Full URLs With Print Styles](css/show-full-urls-with-print-styles.md)
 - [Style A Background With A Linear Gradient](css/style-a-background-with-a-linear-gradient.md)
 - [Using Maps In SCSS](css/using-maps-in-scss.md)
 
@@ -479,11 +483,13 @@ If you've learned something here, support my efforts writing daily TILs by
 ### GitHub
 
 - [Access Your GitHub Profile Photo](github/access-your-github-profile-photo.md)
+- [Create A Private Repo From The CLI](github/create-a-private-repo-from-the-cli.md)
 - [Create And Push To New Repo From CLI](github/create-and-push-to-new-repo-from-cli.md)
 - [List PRs Awaiting Your Review](github/list-prs-awaiting-your-review.md)
 - [Open A PR To An Unforked Repo](github/open-a-pr-to-an-unforked-repo.md)
 - [Open File To Specific Line In Browser](github/open-file-to-specific-line-in-browser.md)
 - [Process JSON Output From gh With jq](github/process-json-output-from-gh-with-jq.md)
+- [Rerun Failed Jobs For Current Branch](github/rerun-failed-jobs-for-current-branch.md)
 - [Target Another Repo When Creating A PR](github/target-another-repo-when-creating-a-pr.md)
 - [Tell gh What The Default Repo Is](github/tell-gh-what-the-default-repo-is.md)
 
@@ -494,6 +500,7 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Disable A Workflow With The gh CLI](github-actions/disable-a-workflow-with-the-gh-cli.md)
 - [Reference An Encrypted Secret In An Action](github-actions/reference-an-encrypted-secret-in-an-action.md)
 - [Run Schedule Action To Commit Regular Updates](github-actions/run-scheduled-action-to-commit-regular-updates.md)
+- [Set Repository Secret From CLI](github-actions/set-repository-secret-from-cli.md)
 - [Trigger A Workflow Via An API Call](github-actions/trigger-a-workflow-via-an-api-call.md)
 - [Use Labels To Block PR Merge](github-actions/use-labels-to-block-pr-merge.md)
 
@@ -748,6 +755,10 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Translate To Canonical Tags With Small Model](llm/translate-to-canonical-tags-with-small-model.md)
 - [Use The llm CLI With Claude Models](llm/use-the-llm-cli-with-claude-models.md)
 
+### Logseq
+
+- [Access Daily Journal From CLI With HTTP Server](logseq/access-daily-journal-from-cli-with-http-server.md)
+
 ### Mac
 
 - [Access All Screen And Video Capture Options](mac/access-all-screen-and-video-capture-options.md)
@@ -1872,6 +1883,7 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Make Neovim The Default Way To View Man Pages](unix/make-neovim-the-default-way-to-view-man-pages.md)
 - [Manually Pass Two Git Files To Delta](unix/manually-pass-two-git-files-to-delta.md)
 - [Map A Domain To localhost](unix/map-a-domain-to-localhost.md)
+- [Merge A Sequence Of Lines Of Output](unix/merge-a-sequence-of-lines-of-output.md)
 - [Move A List Of Files To Another Directory](unix/move-a-list-of-files-to-another-directory.md)
 - [Negative Look-Ahead Search With ripgrep](unix/negative-look-ahead-search-with-ripgrep.md)
 - [Occupy A Local Port With Netcat](unix/occupy-a-local-port-with-netcat.md)
@@ -1892,6 +1904,7 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Rename A Bunch Of Files By Constructing mv Commands](unix/rename-a-bunch-of-files-by-constructing-mv-commands.md)
 - [Repeat Yourself](unix/repeat-yourself.md)
 - [Replace Pattern Across Many Files In A Project](unix/replace-pattern-across-many-files-in-a-project.md)
+- [Restart Process Managed By `process-compose`](unix/restart-process-managed-by-process-compose.md)
 - [Restart Specific Overmind Process](unix/restart-specific-overmind-process.md)
 - [Reverse Each Line Of A File](unix/reverse-each-line-of-a-file.md)
 - [Run A Command Repeatedly Several Times](unix/run-a-command-repeatedly-several-times.md)
@@ -2143,12 +2156,14 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Change Window Name In iTerm](workflow/change-window-name-in-iterm.md)
 - [Configure Email Redirect With Cloudflare](workflow/configure-email-redirect-with-cloudflare.md)
 - [Control Media With Drop Keyboard](workflow/control-media-with-drop-keyboard.md)
+- [Create 1Password Login Item From The CLI](workflow/create-1password-login-item-from-the-cli.md)
 - [Convert An ePub Document To PDF On Mac](workflow/convert-an-epub-document-to-pdf-on-mac.md)
 - [Create A Local Sanity Dataset Backup](workflow/create-a-local-sanity-dataset-backup.md)
 - [Create A Public URL For A Local Server](workflow/create-a-public-url-for-a-local-server.md)
 - [Create Todo Items In Logseq](workflow/create-todo-items-in-logseq.md)
 - [Do Project Time Tracking From The CLI](workflow/do-project-time-tracking-from-the-cli.md)
 - [Enable Dev Tools For Safari](workflow/enable-dev-tools-for-safari.md)
+- [Ensure Tooltips Appear In Screenshots](workflow/ensure-tooltips-appear-in-screenshots.md)
 - [Forward Stripe Events To Local Server](workflow/forward-stripe-events-to-local-server.md)
 - [Get URL For GitHub User Profile Photo](workflow/get-url-for-github-user-profile-photo.md)
 - [Get Your Public IP Address](workflow/get-your-public-ip-address.md)

claude-code/build-from-my-phone-in-cloud-environments.md

@@ -0,0 +1,27 @@
+# Build From My Phone In Cloud Environments
+
+I mostly use Claude Code from a terminal with the `claude` CLI. Everything is
+running on my machine (except the inference). The codebase is already there, the
+feature branch is already in flight, all my tools and dotfiles are installed,
+the dev server or container VMs are running, etc. It's the closest to how I
+already worked before coding harnesses and allows me to seamlessly flow between
+agentic code, review, making my own edits, and so forth.
+
+However, I've been finding it pretty satisfying to open _Claude Code_ either in
+the MacOS app or iOS app, connect a GitHub repo, and then ask for a quick
+feature landed as a PR.
+
+This flow uses [Claude in the Cloud](https://code.claude.com/docs/en/claude-code-on-the-web) to boot up a
+development environment, pull down your codebase, create a branch, build a
+feature, run tests, boot the app, do all the different explorations and checks
+that `claude` typically does, and eventually land a PR back on GitHub.
+
+What I find particularly remarkable about this flow is that this _cloud session_
+is effectively synced across any device where I'm using Claude. So while I'm on
+the train home from the gym, I open the Claude iOS app, pick my repo, and
+describe a feature. By the time I get home, I open Claude.app on my computer
+where I can skim through the chat summary, view the diff, and then suggest next
+steps or jump to the PR on GitHub.
+
+I'm finding this workflow particularly exciting for side projects because I can
+kick off feature experiments from anywhere when the idea comes to me.

claude-code/pick-up-session-locally-that-started-in-cloud.md

@@ -0,0 +1,33 @@
+# Pick Up Session Locally That Started In Cloud
+
+I've been enjoying using _Claude Code_ from the iOS and MacOS apps for various
+side tasks recently. For smaller things I can usually get most of the way there
+from the app. And eventually I pull down the branch or have Claude create a PR
+that I can review.
+
+Sometimes I want to transfer the session onto my computer to pick up where it
+left off for various reasons (e.g. I want the session to have access to some
+local data or specific dependencies).
+
+The `claude` CLI has a flag for that — `--teleport`.
+
+> Resume a teleport session, optionally specify session ID
+
+```
+❯ claude --teleport
+
+Select a session to resume (jbranchaud/still):
+
+      Updated  Session Title
+❯ 1. 21m ago  Blogmark link search modal
+  2. 11h ago  Mobile feed reader margins
+  3. 11h ago  Feed pagination with Pagy
+  4. 1d ago   RSS feed highlights and annotations
+  5. 1d ago   RSS feed categories
+  6. 3d ago   Blogmarks image drag-and-drop upload
+
+↑/↓ to select · Enter to confirm · Esc to cancel
+```
+
+This is similar to how `--resume` works, but it is for a session that was
+previously running in the cloud.

css/show-full-urls-with-print-styles.md

@@ -0,0 +1,46 @@
+# Show Full URLs With Print Styles
+
+Web pages are best experienced in a browser where you can fully interact with
+them. Links, for instance, can be hovered, inspected, and clicked on to navigate
+to the target location -- that's a core part of the web browser experience and
+what makes the internet.
+
+Some people have printers though and they want to be able to print off web
+pages. Any web page can be printed off as is, but if your site or app is
+anticipating certain pages being printed, then you may add some
+`@print`-specific CSS styling to improve how it looks in print format.
+
+Links are a prime example of lost interactivity. When printed, links render as
+blue underlined text with no way to know what they are linking to. One trick
+that I recently picked up from [this lobste.rs thread](https://lobste.rs/s/olnqh2/websites_should_be_printable#c_wigxfy) is to
+add _after content_ that renders the actual, full URL next to the link.
+
+```css
+@media print 
+{
+  a[href]::after
+  {
+    content: " [" attr(href) "]";
+  }
+}
+```
+
+This means that when a page has `<a href="https://still.visualmode.dev">still</a>`, the printed version will not
+just render the word "still", but will be styled to render `still [https://still.visualmode.dev]`.
+
+This will add content after links for both absolute URLs and relative paths (and
+anchor tags). If I wanted to limit the after content to only show absolute URLs,
+I could add a bit more specificity to the matching tag.
+
+```css
+@media print 
+{
+  a[href^="http"]::after
+  {
+    content: " [" attr(href) "]";
+  }
+}
+```
+
+The `^=` matches the pattern of `http` on the start of the `href` string which
+means it will only display absolute URLs.

github-actions/set-repository-secret-from-cli.md

@@ -0,0 +1,31 @@
+# Set Repository Secret From CLI
+
+I set up a GitHub Action that will run daily and potentially push data up to a
+3rd-party API. To be able to post to that API, the action must include a secret
+as the bearer token in the requests. I need to securely include that secret in
+the GitHub Actions environment for this repo.
+
+I can do that from the CLI with `gh` like so:
+
+```bash
+❯ gh secret set TIL_SYNC_TOKEN
+? Paste your secret: ****************************************************************
+```
+
+Notice that it prompts me for the secret and stars out what I paste in. That way
+the secret doesn't end up in my bash history and isn't visible on my screen.
+
+After I've run this, I can see that the secret is set by opening up the repo in
+the GitHub web UI, clicking on _Settings_ > _Secrets and Variables_ > _Actions_.
+I should see `TIL_SYNC_TOKEN` listed under the _Repository Secrets_ section.
+
+Here is an excerpt from my workflow action sync script where that secret is
+referenced -- notice the `${{ secrets.TIL_SYNC_TOKEN }}`.
+
+```yaml
+      - name: Sync posts
+        env:
+          TIL_SYNC_URL: ${{ vars.TIL_SYNC_URL }}
+          TIL_SYNC_TOKEN: ${{ secrets.TIL_SYNC_TOKEN }}
+        run: ruby script/til_sync/til_sync.rb --repo til ${{ inputs.all && '--all' || '--days 3' }}
+```

github/create-a-private-repo-from-the-cli.md

@@ -0,0 +1,34 @@
+# Create A Private Repo From The CLI
+
+The vast majority of the GitHub repos that I've created over my career have been
+done from the web UI. It was only recently that I found out that [I can create a
+new repo from the CLI with `gh`](create-and-push-to-new-repo-from-cli.md).
+
+A repo can be created for the current directory (where a local git repo already
+exists) using the `gh repo create` command. I'm wanting this to be a private
+repo so I can include the `--private` flag instead of the `--public` flag.
+
+```bash
+❯ gh repo create marathon-training --private --source=. --remote=origin --push
+✓ Created repository jbranchaud/marathon-training on github.com
+  https://github.com/jbranchaud/marathon-training
+✓ Added remote https://github.com/jbranchaud/marathon-training.git
+Enumerating objects: 125, done.
+Counting objects: 100% (125/125), done.
+Delta compression using up to 16 threads
+Compressing objects: 100% (114/114), done.
+Writing objects: 100% (125/125), 56.69 KiB | 14.17 MiB/s, done.
+Total 125 (delta 52), reused 0 (delta 0), pack-reused 0 (from 0)
+remote: Resolving deltas: 100% (52/52), done.
+To https://github.com/jbranchaud/marathon-training.git
+ * [new branch]      HEAD -> main
+branch 'main' set up to track 'origin/main' by rebasing.
+✓ Pushed commits to https://github.com/jbranchaud/marathon-training.git
+```
+
+Since I am running this command from the root directory of the git-backed
+project, I specify the `--source` as `.` (mean _here_, the current directory).
+The remaining flags tell it what the remote should be set up as and that I want
+to push this local repo up to the GitHub remote once it is created.
+
+See `gh repo create --help` for more details.

github/rerun-failed-jobs-for-current-branch.md

@@ -0,0 +1,30 @@
+# Rerun Failed Jobs For Current Branch
+
+I have a branch that I've already opened a PR for. I've made more changes, run
+the tests, and pushed up the latest to the remote. I know the tests should have
+been green, but I'm seeing a flickering test fail in CI. I want to retrigger the
+relevant failed jobs to see if I can get the flickers to pass.
+
+First, I need to figure out the workflow ID with any failed jobs.
+
+```bash
+❯ gh run list --branch "$(git branch --show-current)" --status failure
+STATUS  TITLE           WORKFLOW      BRANCH        EVENT         ID           ELAPSED  AGE
+X       Better comp...  Ruby on R...  jbranchau...  pull_request  38778306948  10m54s   about 25 ...
+```
+
+This one-liner does that by referencing the current branch (via `git`) and then
+asking `gh` to list failed workflows associated with that branch.
+
+Grabbing the `ID` value, I can then ask `gh` to rerun failed jobs for that
+workflow.
+
+```bash
+❯ gh run rerun 38778306948 --failed
+✓ Requested rerun (failed jobs) of run 38778306948
+```
+
+Opening that PR up in the browser, I can see that the specific workflow has been
+rerun and is currently _in progress_.
+
+See `gh run list --help` and `gh run rerun --help` for more details.

logseq/access-daily-journal-from-cli-with-http-server.md

@@ -0,0 +1,72 @@
+# Access Daily Journal From CLI With HTTP Server
+
+I use [logseq](https://logseq.com/) as the daily driver for my notes both inside
+and outside of work. I like the organization provided by the outliner notetaking
+style.
+
+I now want to set up some automations that interact with my notes. I originally
+figured this would involve reading directly from markdown files on disk.
+However, they recently migrated to a SQLite storage format and I learned that
+there is an HTTP server you can run for more reliable interactions.
+
+The HTTP server needs to be enabled. To do this, I did the following (though I'm
+not sure if _Developer Mode_ is required):
+
+1. Find the `...` icon to access the app's main menu
+2. Open _Settings_
+3. Enable _Developer Mode_ under _Advanced_
+4. Enable _HTTP API Server_ under _Features_
+5. Restart the logseq app
+6. Click the disabled _API_ icon that is in the upper-right next to the
+   navigational arrows
+7. Create a new _Authorization Token_ and copy the value
+8. Click _Start Server_ and it will run at `http://127.0.0.1:12315`
+
+With all that in place, the server is running and ready to accept my first
+query. I want to get a look at my notes so far today, so I'll run a `curl`
+request against the `/api` path with the auth token I copied.
+
+```bash
+❯ curl -s -X POST http://127.0.0.1:12315/api \
+    -H "Authorization: Bearer $LOGSEQ_API_TOKEN" \
+    -H "Content-Type: application/json" \
+    -d '{"method":"logseq.Editor.getPageBlocksTree","args":["Oct 31st, 2026"]}'
+[{"updatedAt":1793456696397,"createdAt":1793456670024,"parent":{"id":5943},"children":[],"id":5947,"level":1,"uuid":"6ae5fa1e-afa2-46c9-abbc-abf58b434e3e","fullTitle":"Working on setting up the HTTP server for logseq to see if I can interact with it via the CLI.","content":"Working on setting up the HTTP server for logseq to see if I can interact with it via the CLI.","order":"a0","title":"Working on setting up the HTTP server for logseq to see if I can interact with it via the CLI.","page":{"id":5943}},{"updatedAt":1793456702190,"createdAt":1793456697156,"parent":{"id":5943},"children":[{"updatedAt":1793456702212,"createdAt":1793456702212,"parent":{"id":5948},"children":[],"id":5949,"level":2,"uuid":"6ae5fa3e-0865-4713-9690-c456fb1a37ad","fullTitle":"","content":"","order":"a0","title":"","page":{"id":5943}}],"id":5948,"level":1,"uuid":"6ae5fa39-96bc-4f26-9d5f-2b3b3a366070","fullTitle":"[[6aa01dba-1b28-4540-b205-5254ee37d825]]","content":"[[6aa01dba-1b28-4540-b205-5254ee37d825]]","order":"a1","title":"[[6aa01dba-1b28-4540-b205-5254ee37d825]]","page":{"id":5943},"refs":[{"id":642}]}]%
+```
+
+The data we pass as the request body includes the `method` we are running which
+is "get all the blocks for the specified page"
+([`getPageBlocksTree`](https://plugins-doc.logseq.com/logseq/Editor/getPageBlocksTree)).
+The `args` then specify what page we are referring to. In this case, today's
+daily journal -- `"Oct 31st, 2026"` (👻). The result is that pile of JSON.
+
+Whenever I get a long line of JSON like that, I like to use `jq` at the very
+least to pretty-print it or even to dig in a bit.
+
+```bash
+❯ curl -s -X POST http://127.0.0.1:12315/api \
+    -H "Authorization: Bearer $LOGSEQ_API_TOKEN" \
+    -H "Content-Type: application/json" \
+    -d '{"method":"logseq.Editor.getPageBlocksTree","args":["Oct 31st, 2026"]}' | jq
+[
+  {
+    "updatedAt": 1793456696397,
+    "createdAt": 1793456670024,
+    "parent": {
+      "id": 5943
+    },
+    "children": [],
+    "id": 5947,
+    "level": 1,
+    "uuid": "6ae5fa1e-afa2-46c9-abbc-abf58b434e3e",
+    "fullTitle": "Working on setting up the HTTP server for logseq to see if I can interact with it via the CLI.",
+    "content": "Working on setting up the HTTP server for logseq to see if I can interact with it via the CLI.",
+    "order": "a0",
+    "title": "Working on setting up the HTTP server for logseq to see if I can interact with it via the CLI.",
+    "page": {
+      "id": 5943
+    }
+  },
+  ...
+]
+```

unix/merge-a-sequence-of-lines-of-output.md

@@ -0,0 +1,42 @@
+# Merge A Sequence Of Lines Of Output
+
+More often than not when I'm chaining a sequence of Unix commands together, I'm
+eventually trying to fan them out or execute each argument separately. I
+recently had the need for a bunch of separate lines of output to be fanned back
+in to a single line.
+
+The tool I found for this job is `paste` (not the most intuitive name).
+
+> merge corresponding or subsequent lines of files
+
+I was listing out all the categories that my active [TIL posts](https://github.com/jbranchaud/til) full under. I used the following
+command to do that.
+
+```bash
+❯ git ls-files '*/*.md' | cut -d/ -f1 | sort -u
+ack
+ansible
+astro
+aws
+bash
+brew
+chrome
+...
+```
+
+Those are all my categories, but then the format I really wanted was for them to
+be all on the same line with a clear delimiter so that I could paste them for
+[use in another script](/llm/translate-to-canonical-tags-with-small-model.md).
+
+The trick is to tack on one more pipe to `paste`. The `-s` tells `paste` to join
+every line from a given input file (e.g. stdin) onto a single line. The `-d`
+allows for overriding the default delimiter (tab) with something like a `,`.
+Lastly, the trailing `-` tells `paste` to read from stdin rather than some other
+named file(s).
+
+```bash
+❯ git ls-files '*/*.md' | cut -d/ -f1 | sort -u | paste -sd, -
+ack,ansible,astro,aws,bash,brew,chrome,claude-code,clojure,css,cursor,deno,devops,docker,drizzle,elixir,gatsby,git,github,github-actions,go,groq,heroku,html,http,inngest,internet,java,javascript,jj,jq,kitty,linux,llm,mac,math,mise,mongodb,mysql,neovim,netlify,next-auth,nextjs,phoenix,planetscale,pnpm,postgres,prisma,python,rails,react,react_native,react-testing-library,reason,remix,rspec,ruby,sed,shell,sqlite,streaming,tailwind,taskfile,tmux,typescript,unix,vercel,vim,vscode,webpack,workflow,xstate,yaml,zed,zod,zsh
+```
+
+See `man paste` for more details.

unix/restart-process-managed-by-process-compose.md

@@ -0,0 +1,28 @@
+# Restart Process Managed By `process-compose`
+
+I'm working on a project where we are experimenting with
+[`process-compose`](https://github.com/F1bonacc1/process-compose) as a better
+way orchestrate a set of processes that all need to be running for a web
+development server (Rails) to work. This includes the Rails server itself, a
+JavaScript build process, PostgreSQL, a background worker process, etc.
+
+In the past [I've used `overmind` with
+Procfiles](/rails/run-dev-processes-with-overmind-instead-of-foreman.md) to do a
+similar sort of thing. No matter what tooling is in play, from time to time, I
+will eventually need to restart one of the processes that has become stale.
+
+The `vite` (JavaScript build) process is a good example of this. Configuration
+changes, new `npm` dependencies, and tailwind classes on new files can all lead
+to the build being stale.
+
+From the `process-compose` TUI, I can navigate the cursor over the `vite`
+process and hit `Ctrl-r` to restart the process.
+
+I can also issue a restart via the CLI like so:
+
+```
+❯ mise exec -- process-compose process restart vite
+Process vite restarted
+```
+
+[source](https://f1bonacc1.github.io/process-compose/cli/process-compose_process_restart/)

workflow/create-1password-login-item-from-the-cli.md

@@ -0,0 +1,38 @@
+# Create 1Password Login Item From The CLI
+
+The default shape of a login item in 1Password is _Username_ and _Password_.
+Whenever I'm trying to save something in 1Password, like an AWS Access Key for
+example, I find myself manually adjusting the fields to get it into the right
+shape for that item.
+
+What I realized would be nicer is using [`op` (the 1Password
+CLI)](https://www.1password.dev/cli) to create a login item with the shape I
+want (which I can later reference in my shell history). Then I can pop open the
+desktop app and update the latest entry with the actually secret values that I
+copy from, in this case, the AWS Console.
+
+```bash
+❯ op item create --vault "Personal" --category login --title "AWS - jbranchaud-lastword-cli" \
+  'Access Key ID[text]=TODO' \
+  'Secret Access Key[concealed]=TODO' \
+  'Default region[text]=us-east-1'
+ID:          abc123
+Title:       AWS - jbranchaud-lastword-cli
+Vault:       Personal (efg456)
+Created:     now
+Updated:     now
+Favorite:    false
+Version:     1
+Category:    LOGIN
+Fields:
+  Access Key ID:        TODO
+  Secret Access Key:    [use 'op item get abc123 --reveal' to reveal]
+  Default region:       us-east-1
+```
+
+In this command I point to my `Personal` vault, I update the title of the login
+item inline, and then I have some custom fields on the item that match the shape
+of this scenario.
+
+When I hit enter, a listing of fields is output. I can then hop over to the
+desktop app to set the actually ID and secret for the access key.

workflow/ensure-tooltips-appear-in-screenshots.md

@@ -0,0 +1,18 @@
+# Ensure Tooltips Appear In Screenshots
+
+I use [CleanShot X](https://cleanshot.com/) for all my screenshot needs. It's by
+far the best tool for the job.
+
+I'm a big proponent of taking high-context screenshots. That is, including
+everything in the screenshot that another person viewing it would need to know
+so that they don't have to answer follow up questions. This usually means other
+parts of the window besides the focal point or including the URL bar.
+
+Another, less common, aspect of a shot I sometimes need, especially in a
+browser, are hover-based visual elements. That could be a tooltip or the option
+in a dropdown the cursor is over. Sometimes it is even a visual element that
+goes away after a beat.
+
+To capture these with CleanShot X, I need to turn on the "Freeze screen when
+taking a screenshot" setting which can be found in _Settings_ under the
+_Screenshots_ tab in the _Capture_ section.