Growing Self Workflow Plan

Where Each Job Stands

Eight workflows are built and tested on past episodes, past page briefs and practice info sessions. One is in the team library today. The other seven need a pass on Lisa's voice and a tested guide before anyone runs them. Three more are new. Sinatra and Rocket came out of the October 30 call and the QC rubric, and Zeppelin waits for Mandi's meeting.

The Two Studios

This is the planned setup. Neither studio is built yet.

Each studio is one shared Claude project with its own instructions, a short list of knowledge files and the skills it runs. Say hi and it answers with a menu. Every job has a named team member, so it is easy to remember who does what, and each one hands off to the next by name. The skills live in the team library, so either studio can call any of them.

From Guest Booking to Newsletter

One episode moves left to right. After Eddie writes the card, three jobs run side by side, and the week's episodes feed Lettie's newsletter. Rocket closes the loop every Monday, so what worked last week changes how Preppy and Eddie work next week.

• Handoff to the next job
• Feedback loop
• A named team member in the studio

Every Workflow

Pick a team member to see their workflow. Each one shows who runs it, what starts it, the skill it uses, what comes back, who approves it, where it lands, and what is left. Scores are out of 2. A skill passes at 1.6 or more on every line of its rubric.

LHS Podcast Studio

Preppy LongstockingEn-Titled EddieAntie SocialBlog MarleyEwe-TubeLettie RipThank SinatraRocket the Raccoon

Growth Studio

Wrap StarPaige TurnerLead Zeppelin

Guest Prep Doc

Preppy Longstocking in LHS Podcast Studio

Starts whenGuest booked

Runs itPreppy Longstocking

Hands backPrep page and sources sheet

ApprovesHayley checks, Lisa reads

Lands inDrive and production board

Who runs it

Hayley

Trigger and input

A guest is booked. Guest name, show (LHS or LHSFT), planned topic, and the guest's pitch or links.

Skill

guest-prep-doc (02)

Output

One page for Lisa (the angle, a question flow in her five interview movements, Cloudprompter lines) and a sources sheet for Hayley with every fact tied to its page.

Approver

Hayley checks the sources. Lisa reads it before recording.

Lands in

Google Drive and the episode record on the production board.

Replaces

Content Planning Command Center.

Status

Built, tuning 1.79 of 2 after 5 versions. Under the 1.6 line on Lisa's voice (1.38), ready to use (1.40) and fits the place (1.50).

What is left

• Compare it line by line with the Content Planning Command Center instructions and the LHS Media QC rubric, and add the rules it is missing
• Hayley decides whether the 30-second hook, intro, outro and clip plan move into the prep doc
• Hayley decides whether sponsor points belong here or stay on the YAP Promos board
• Publish it to the team library, then a step-by-step guide tested from Hayley's seat

Episode Title and Description

En-Titled Eddie in LHS Podcast Studio

Starts whenFinal transcript

Runs itEn-Titled Eddie

Hands backEpisode card

ApprovesHayley approves

Lands inMegaphone

Who runs it

Hayley

Trigger and input

Mark's final edit is done. The timestamped transcript, plus the prep doc if there was one.

Skill

episode-title-description (01), the Episode Card

Output

An episode card with three search-tested titles and the reason for each, the description in Lisa's voice, the intro clip and line, Megaphone tags and the file name. It checks the episode against the prep doc and asks for a fresh Semrush check when the topic moved.

Approver

Hayley. Lisa marks titles during calibration. Who signs off after that is an open decision.

Lands in

Megaphone, by hand.

Replaces

Transcript Reality Check, LHS Podcast Descriptions, LHSFT Podcast Descriptions, and the ChatGPT rewrite step.

Status

In rebuild Version 1 is in the team library since October 30. Its last full test, on 6 episodes, passed every line except hook strength (1.50). A targeted rerun on 3 episodes then scored 1.91 of 2. A version rebuilt to Hayley's current standard waits on her four checks.

What is left

• The rebuilt version follows Hayley's current standard and waits on her answers to four checks. The "| Love | E571" title tail, the extra growingself.com address in the call to action, no sign-off, and her 150-word example against her 300 to 500 word rule
• Calibration. Max sends Lisa a title test with 10 past episodes and 3 titles for each. Lisa marks each yes or no, with a reason
• Hayley lists the titles Lisa approved and the ones she turned down
• Lisa uploads the rebuilt version to the team library. Version 1 stays available until then

Episode Social Set

Antie Social in LHS Podcast Studio

Starts whenFinal transcript

Runs itAntie Social

Hands backSocial set

ApprovesHayley approves

Lands inHannah builds and posts

Who runs it

Hayley

Trigger and input

The final transcript is ready. The transcript and the episode card.

Skill

episode-social-set (03)

Output

For LHS, an Instagram set with clip picks at real timestamps, the reel caption, carousel slides and caption, the quote card, the story, a new DM keyword with its reply, and the newsletter blurb. For LHSFT, a LinkedIn set with clip picks, a text post with its cover line, a cutdown clip caption, and three LinkedIn board items. Both come with Hannah's board items and tasks as paste-ready text.

Approver

Hayley, then Hannah builds from it.

Lands in

Hannah's design queue, then Instagram for LHS and LinkedIn for LHSFT, by hand.

Replaces

Part of LHS Media Promotional Content.

Status

Built, tuning 1.73 of 2 after 5 versions. Under the line on faithful (1.5), Lisa's voice (1.25) and timestamps (1.5).

What is left

• Rerun with the 30 Instagram captions and 30 LinkedIn posts in Hayley's examples
• Compare it with the LHS Media Promotional Content project files
• Guide for Hayley and Hannah, tested from their seats

Episode Blog Brief

Blog Marley in LHS Podcast Studio

Starts whenVideo is up

Runs itBlog Marley

Hands backBlog brief

ApprovesHayley approves

Lands inMat publishes

Who runs it

Hayley

Trigger and input

The episode video is up. The episode, its video and the episode card.

Skill

episode-blog-brief (04)

Output

One brief Mat's builder can publish without bouncing it back, with every field in his Airtable order, chapter times checked against the video, live links only, and a first line that says whether it is ready for Mat.

Approver

Hayley, then Mat publishes.

Lands in

WordPress, built by Mat.

Replaces

Content Production Command Center (blog half).

Status

Built, tuning 1.71 of 2 after 5 versions. Under the line on faithful, Lisa's voice, safe, fits the place, and work left for Hayley.

What is left

• Hayley shares the Content Production Command Center project. It is the one project Max's account cannot see
• Lisa decides how spoken "free consultation" in transcripts is handled. Mat's guide bans the phrase, and until she decides, every brief that contains it holds
• Lisa confirms which guide wins, the May 2026 build guide or the older Requirements Guide
• Mat confirms his builder reads chapter times written as [MM:SS]
• Compare the skill with that project's rules and the QC rubric's SEO section
• Guide for Hayley and Mat

YouTube Title and Description

Ewe-Tube in LHS Podcast Studio

Starts whenPrivate upload

Runs itEwe-Tube

Hands backUpload sheet

ApprovesIndia approves

Lands inYouTube Studio

Who runs it

India

Trigger and input

The edited video is uploaded as private. The private upload's captions or final-cut transcript.

Skill

youtube-title-description (06)

Output

One upload sheet with three titles (search, browse, second search), a paste-ready description with checked chapters, titles and times for the three cutdowns, and thumbnail text with an alternate.

Approver

India.

Lands in

YouTube Studio, by hand.

Replaces

Content Production Command Center (YouTube half).

Status

Built, tuning 1.75 of 2 after 5 versions. Under the line on Lisa's voice (1.38) and chapters true (1.38).

What is left

• India gives Max access to YouTube Studio so the skill can be tested on real uploads
• Guide for India, tested from her seat

Weekly Newsletter

Lettie Rip in LHS Podcast Studio

Starts whenWeekly, one line

Runs itLettie Rip

Hands backNewsletter draft

ApprovesHayley and Lisa

Lands inJuven sends

Who runs it

Hayley

Trigger and input

Once a week. One line with the show (LHS or LHSFT) and the send date.

Skill

weekly-newsletter (05)

Output

Three subject lines with reasons, preview text, the paste-ready newsletter in Lisa's voice built from that week's episodes, and every link checked.

Approver

Hayley, with Lisa on voice. The newsletter shape is an open decision.

Lands in

Pipedrive for LHS, Kajabi for LHSFT. Juven sends.

Replaces

No current tool. Hayley writes it by hand.

Status

Built, tuning 1.83 of 2 after 5 versions. Under the line on Lisa's voice (1.33).

What is left

• Lisa picks the shape. Her real LHS letters run 1,000 to 1,260 words with LOVE, HAPPINESS, SUCCESS and HEARTBREAK sections and end "With love,". The skill writes a 200 to 450 word digest today
• Lisa decides whether words like "unpack" and "you deserve" stay allowed, since her letters use them
• Guide for Hayley and Juven

Thank-You and Guest Follow-Up

Thank Sinatra in LHS Podcast Studio

Starts whenRecording done

Runs itThank Sinatra

Hands backThank-you and follow-up

ApprovesLisa reviews

Lands inSent by hand

Who runs it

Hayley

Trigger and input

Recording is done. The guest, one moment from the conversation, and the release date.

Skill

None yet

Output

A thank-you note in Lisa's voice that names a moment from the talk, the release-date note for the guest's team, and the sharing ask.

Approver

Lisa reviews and personalizes the note.

Lands in

Email from Lisa or Hayley, by hand.

Replaces

Section 8 of the QC rubric, which no skill covers today.

Status

Planned Not built. Proposed because the QC rubric asks for it within 48 hours of every recording.

What is left

• Hayley says whether she wants it
• If yes, it is built and tested the same way as the others

Monday Numbers

Rocket the Raccoon in LHS Podcast Studio

Starts whenMonday 9am

Runs itRocket the Raccoon

Hands backMonday note

ApprovesHayley approves changes

Lands inSkill rules updated

Who runs it

Runs on its own, on the schedule Hayley picks

Trigger and input

A schedule, for example Monday at 9:00am ET. Last week's numbers from the sources Hayley picks, and the QC rubric testing notes.

Skill

None yet (a scheduled task)

Output

One note for Hayley with what sat below the usual or below the benchmark, what beat the usual and why, and one protocol change to try.

Approver

Hayley approves each protocol change. Who owns content results is the team's Monday decision.

Lands in

The workflow tracker in Notion, and a ping to Hayley.

Replaces

Nothing. Today nobody reviews the numbers on a schedule.

Status

Planned Not built. Mandi asked for it on the October 30 call. Plan in the Monday Numbers section.

What is left

• Hayley picks the day, the time and the metrics
• Connect each source, or set up a weekly export where no connector exists. Semrush is connected today
• Collect a benchmark for each metric, each with its source and date
• The team decides who owns content results

Info-Session Recap

Wrap Star in Growth Studio

Starts whenSession ends

Runs itWrap Star

Hands backRecap package

ApprovesLisa approves

Lands inMandi sends

Who runs it

Mandi

Trigger and input

A therapist info session ends. The session notes or transcript.

Skill

info-session-recap (07)

Output

The recap email from Lisa, one next step with a working link, an answer to every question asked, notes for each attendee's follow-up, and every fact tied to its public source.

Approver

Lisa, since the email goes out in her name.

Lands in

Sent by Mandi, from Kajabi or Gmail (to decide).

Replaces

Part of CCFT Command Center.

Status

Built, tuning The last full scoring (version 5, on 7 practice sessions) averaged 1.8 of 2, under the line on Lisa's voice (1.14). After the audit fixes, a targeted rerun on 4 sessions scored lower on faithful (1.25), Lisa's voice (1.00) and ready to use (1.50). It has not had a fresh full scoring since, and has not run on a real session.

What is left

• Mandi fills in tuition, dates, deposit and the enrollment link
• Lisa says whether a session transcript may be used, since the session page says it is not recorded
• Mandi exports two or three real recaps from Cersei for the voice pass
• Mandi picks Kajabi or Gmail for the send

Landing Page Brief

Paige Turner in Growth Studio

Starts whenNext page on the list

Runs itPaige Turner

Hands backPage brief

ApprovesLisa and Mandi

Lands inMat builds

Who runs it

Mandi

Trigger and input

A page is next on the site list. One outcome, for example Relationship Clarity.

Skill

landing-page-brief (08), updated October 27

Output

A sourced, checked page brief with the H1 from measured search phrases, two honest paths, a waitlist card where no group program exists yet, FAQs, form fields, and yes or no questions for Lisa.

Approver

Lisa answers the yes or no questions. Mandi approves the brief.

Lands in

The new site, built by Mat.

Replaces

The StoryBrand copywriter GPT.

Status

Built, tuning Follows all nine of Lisa's October 27 answers. The last fresh brief passed its checker, but that run was not rescored. Scored versions sat under the line on Lisa's voice (1.00), ready to use (1.00) and approvable (1.38).

What is left

• Lisa answers the 8 open questions in the Relationship Clarity brief
• Lisa confirms the 15-page list, the Healthy Attachment merge and Long Distance Love
• Lisa sets up the page form with Mandi and Kalani, and checks the privacy and security of the database it feeds before any therapy inquiry form goes live
• Lisa's Semrush key, so search phrases can be measured outside Claude

Lead Magnets

Lead Zeppelin in Growth Studio

Starts whenTo decide

Runs itLead Zeppelin

Hands backTo decide

ApprovesMandi

Lands inTo decide

Who runs it

Mandi

Trigger and input

Set at next week's lead magnets meeting. To decide.

Skill

None yet

Output

To decide.

Approver

Mandi, with Lisa.

Lands in

To decide.

Replaces

Nothing yet.

Status

Ideas only Mandi's ideas are laid out in their own section. Nothing is chosen.

What is left

• Mandi and Max meet next week and pick which ideas go first
• Then this row gets a trigger, an input, an output and a skill

Rocket's Monday Numbers

Hayley's QC dashboard, on a schedule

Mandi asked for a loop that taps Hayley on the shoulder on a schedule, so the content gets better from its own numbers and Hayley never has to start that work herself. Rocket is that loop. He is a scheduled task in Claude that runs while Hayley's computer is on, or in the cloud.

Example layout. Numbers are made up

Rocket Monday, 9:00am ET

Below the Usual

Episode A downloads at 7 days ran 22% under the show's 90-day median. Its top title led with the search phrase and scored 6 of 10 on hook.

A weak hook costs listens in the first week, when most listens happen.

Beat the Usual

Episode B's carousel earned 31% more saves than usual. It opened on a question in Lisa's own words.

Saves keep a post in front of people for longer, so this opening is worth repeating.

Change to Try

Eddie. When the search title scores under 7 on hook, show the hook title first.

Hayley approves it or says not now. Rocket never changes a skill by himself.

How Rocket Decides What to Flag

• He compares each number with the show's own usual first, the median of the last 90 days.
• An industry benchmark counts only with its source and date written next to it.
• Each metric has a good direction, set once at setup. Fewer unsubscribes, a lower search position number and fewer hours per episode are good news.
• He flags a number when it moves the wrong way from the usual by more than the size Hayley sets (for example 20%), or two weeks running.
• Every number he flags says what it costs or earns.
• He proposes one change at a time. Each change Hayley approves gets a new skill version and a line in the tracker.
• He reads Section 10 of Hayley's QC rubric, the testing notes, and counts what keeps going wrong.

Metrics Hayley Can Pick

Hayley picks up to five to start. Mandi's September summary puts email clicks near 0.1% while opens sit in the upper 20s, so the click is a likely first pick.

Where the QC Rubric Lives

Hayley's Content Prep QC Rubric has ten sections. This is where each one lives in the studio, and what no skill covers yet.

Setting Rocket Up

1. Hayley picks the day and time. Monday at 9:00am ET lands before the Monday alignment session.
2. Hayley picks up to five metrics from the list above.
3. Each source gets a connector, or a weekly export into one Drive folder where no connector exists. Production numbers come from a new Workflow Runs log in Notion, one row each time a workflow runs, with the episode, the skill version, the date, minutes spent and the rubric sections missed.
4. Each metric gets its usual (the 90-day median) and, where one exists, a benchmark with its source and date.
5. The first two Mondays are a dry run. Rocket writes the note, and Hayley checks it against what she sees by hand.
6. Then it runs on its own. The note lands in the workflow tracker in Notion, with a ping to Hayley.

Mandi's Lead Magnet Ideas

Nothing here is chosen. Mandi and Max pick at next week's lead magnets meeting, and the chosen ideas become Lead Zeppelin's jobs. The numbers below are Mandi's own estimates.

Mandi named the tone on October 30 as Joy Marketing, the humor-led voice that made the Therapist Life swag a hit at the APA booth. In her email tests, negative angles about insurance clawbacks and license risk converted better than hopeful ones. Both belong in the meeting, and every clinical or legal claim goes past Lisa first.

The Notion Version

The same plan, kept in HomeBase so the team can find each workflow, its skill and its diagram in one place. The skills also stay in the Claude team library, so there are two copies. Nothing here is set up yet. It is a proposed HomeBase setup, and Lisa picks where it lives.

Growing Self Workflows
├── Workflows (the tracker, one row per workflow)
├── Workflow Runs (one row per run, for the QC numbers)
├── LHS Podcast Studio (greeting, menu, team, rules)
├── Growth Studio (greeting, menu, team, rules)
├── Rocket's Monday Numbers (metrics, rubric map, setup)
├── Mandi's Lead Magnet Ideas (not decided)
└── Decisions Waiting

The Workflow Tracker

One row per workflow, 11 rows to start. The Skill column links to the Growing Self Skills database already in Lakshmi. Each row's page holds its flow chart, drawn in Notion from a short Mermaid block.

Decisions Waiting

These are the choices only the Growing Self team can make. Each one unblocks a workflow.