Step 1 of 13
💬
The conversation will appear here as you step through →
Y
You

Ask the Automator to watch this project folder. Whenever a new file is added — an email, a meeting note, a report, anything — read it and update project_status.md to reflect what it tells us. Keep these sections current: What's done, What's in progress, Blockers, Decisions made, What's next. Don't just append — rewrite each section so it always reflects the current state, not a log of events.

A
AI Assistant Majk Rabbit

Finding the right automation tool and wiring this up.

🔍 search_skills "automator file watcher folder monitoring"
enable_skill Automator Expert — event-driven automation, file-system watching
👥 getAvailableTeammates finding Majk Rabbit to assign file-processing tasks to
createAutomation "Project Status Updater" — file-system watch on project folder, fires on new file created
A
AI Assistant Majk Rabbit

Done — the automation is live. Here's what's now running:

SettingValue
Watchesproject_alpha/ (all subdirectories)
TriggerAny new file created — email, meeting note, report, anything
ActionRead new file + read project_status.md → rewrite all five sections
RuleDon't append — rewrite each section to reflect current state, not a log

The next time a file lands in the folder, this fires automatically. You don't need to do anything.

Y
You

A new file has been added to the project folder. Your job is to read it and update project_status.md so it reflects the current state of the project.

STEP 1: Read the new file
Read the file at the path provided in the event data. It could be an email, meeting note, report, status update, or any other document. Extract everything relevant: decisions made, work completed, work in progress, blockers, and next steps.

STEP 2: Read the current project_status.md
Read project_status.md if it exists. Understand what is already recorded there.

STEP 3: Rewrite project_status.md
Rewrite (not append to) project_status.md so it always reflects the CURRENT state...

STEP 4: Confirm
Report back: what file was processed, what key information was extracted, and what changed.

Automation Details:
File Path: project_alpha/emails/thread_apr_22.txt
File Name: thread_apr_22.txt
File Extension: .txt
Event Type: create
Triggered at: 2026-04-22T14:47:00.000Z

A
AI Assistant Majk Rabbit

Reading the new file and the current status doc, then rewriting the affected sections.

📄 Read thread_apr_22.txt — email thread, scanning for decisions, blockers, updates
📄 Read project_status.md — reading current state before updating
Write project_status.md — rewriting affected sections based on new information
A
AI Assistant Majk Rabbit

project_status.md updated. Here's what changed:

SectionWhat changed
BlockersAdded: API dependency from Platform team not ready — blocks the data sync feature. ETA unknown.
Decisions madeAdded: Team agreed to build a mock data layer to unblock frontend work while waiting on API.
In progressUpdated: "Data sync feature" now marked as blocked pending API. "Frontend build" updated to note mock data layer approach.
What's nextAdded: Assign someone to own the mock data layer this week.

Everything else in the status doc is unchanged — only the sections affected by this email were rewritten.

Y
You

Ask the Automator to make it so that every Friday at 5pm, Majk Rabbit reads everything that arrived in the project folder this week and produces a polished HTML status report — something I could send to stakeholders. Call it status_report_[date].html. Same sections as the status doc, but designed to be read by people outside the team who need the big picture, not the detail.

A
AI Assistant Majk Rabbit

Setting up the weekly report automation now.

🔍 search_skills "automator scheduled trigger cron weekly"
enable_skill Automator Expert — event-driven automation, scheduled triggers
👥 getAvailableTeammates finding Majk Rabbit to assign the report task to
createAutomation "Weekly Status Report" — schedule: Friday 5:00 PM, reads project_alpha/ and writes status_report_[date].html
A
AI Assistant Majk Rabbit

Done — the weekly report automation is live:

SettingValue
ScheduleEvery Friday at 5:00 PM
ReadsEverything in project_alpha/ that arrived during the week
Outputstatus_report_[date].html — stakeholder-ready, big picture not detail
SectionsSame as the status doc, written for people outside the team

First run: this Friday at 5pm. You don't need to do anything.

Y
You

It's Friday 5:00 PM. Read everything that arrived in project_alpha/ this week and produce a polished HTML status report for stakeholders.

Output file: status_report_[date].html — write it to the project folder. Same sections as the internal status doc, but written for people outside the team who need the big picture, not the detail.

Automation Details:
Automation: Weekly Status Report
Schedule: Friday 5:00 PM
Triggered at: 2026-04-19T17:00:00.000Z

A
AI Assistant Majk Rabbit

Reading everything that arrived in the project folder this week, then building the stakeholder report.

📄 Read thread_apr_17.txt — 3 emails, extracting decisions and action items
📄 Read standup_apr_18.md — daily standup notes
📄 Read thread_apr_22.txt — blocker email, already captured in status doc
📄 Read retro_apr_19.md — sprint retro, extracting wins and concerns
📄 Read project_status.md — using current state as the baseline
Write status_report_apr_19.html — building stakeholder-ready page
A
AI Assistant Majk Rabbit

Done — status_report_apr_19.html is in the project folder. It covers what shipped this week, what's in progress, the active blocker and its workaround, and next week's priorities.

You can find it in project_alpha/status_report_apr_19.html, or ask me to open it.

Y
You

Open the report.

status_report_apr_19.html
📋 Project Alpha — Week of Apr 15–19
Status report generated Friday 5:00 PM · For distribution to stakeholders
6
Items completed this week
1
Active blocker
On track
Overall status
What Shipped This Week
Authentication flow — complete, in review
Dashboard UI components — complete, merged to main
API schema finalised and documented
In Progress
Frontend
72%
Data sync
30%
Testing
15%
⚠️ Active Blocker
  • API dependency from Platform team not ready — blocking data sync feature. Workaround: mock data layer in progress, frontend work continues unblocked.
Next Week
Complete mock data layer — assign owner Monday
Auth flow review complete, merge by Wednesday
Begin integration testing once mock layer is ready
Y
You

Ask the Automator to watch this project folder. Whenever a new file arrives, read it and keep me accurately informed about where this project actually stands. You decide what structure and sections would be most useful — I just want to be able to read one file and know what's happening.

A
AI Assistant Majk Rabbit

Setting up the automation — open goal, structure decided at run time.

🔍 search_skills "automator file watcher folder monitoring"
enable_skill Automator Expert — event-driven automation, file-system watching
👥 getAvailableTeammates finding Majk Rabbit to assign file-processing tasks to
createAutomation "Project Status — Outcome" — file-system watch on project folder, fires on new file created
A
AI Assistant Majk Rabbit

Done — the automation is live:

SettingValue
Watchesproject_alpha/ (all subdirectories)
TriggerAny new file created
GoalKeep you accurately informed — structure decided by the agent each run based on what the project needs

No sections specified. Each time it runs, it'll read the project first and decide what structure would be most useful.

Y
You

A new file has been added to the project folder. Keep the project owner accurately informed about where this project stands.

Goal: Read the new file and update project_status.md so it reflects the current state. You decide what structure and sections are most useful for this specific project — don't just fill in a template.

Automation Details:
File Path: project_alpha/emails/thread_apr_22.txt
File Name: thread_apr_22.txt
Event Type: create
Triggered at: 2026-04-22T14:51:00.000Z

A
AI Assistant Majk Rabbit

Reading the full project folder first to understand the context before deciding what structure would be most useful.

📄 Read emails/ — reading all saved threads to understand communication patterns
📄 Read meetings/ — reading standup and retro notes
📄 Read reports/ — reading weekly reports for context
📄 Read thread_apr_22.txt — new file triggering this run
Write project_status.md — building structure based on what this project actually needs
A
AI Assistant Majk Rabbit

Here's how the two approaches compare on the same folder:

Method promptingOutcome prompting
StructureYour 5 sections, exactly as specified6 sections — added one you didn't ask for
Added section"Risk register" — the email threads show a pattern of late-breaking dependencies; this project needs a dedicated risk section
BlockersListed as a sectionPromoted to the top — given the API dependency, it's the most important thing to see first
What you getPredictable, easy to skim, you know where everything isStructured around what this project actually needs to surface

The risk register wouldn't have appeared in the method version because you didn't ask for it. But after reading the email patterns, it's clearly what this project needs most.

Y
You

Ask the Automator for a list of automations — which are enabled, when they last ran, and how to pause or change one.

A
AI Assistant Majk Rabbit

Here are your active automations:

NameTriggerStatusLast ran
Project Status UpdaterNew file in project_alpha/✅ Active2:47 PM today
Weekly Status ReportFriday 5:00 PM✅ ActiveLast Friday

To pause, change the trigger, or delete a rule — just say which one and what you'd like to do. You can also see all automations and their full run history in Apps → Automator.

Y
You

If a new file mentions a blocker or missed deadline, flag it separately so I notice it immediately.

A
AI Assistant Majk Rabbit

Added to the rule. Now whenever a new file triggers an update, I'll also check for blockers or missed deadlines. If I find one, I'll write a separate ALERT.md in the project root — one file, always the current alert, overwritten each time. If there's nothing to flag, the file stays empty.

The status doc update still runs regardless — the alert is an addition, not a replacement.

Y
You

Each time the status updates, append a one-line entry to a changelog so I can see what changed and when.

A
AI Assistant Majk Rabbit

Added to the rule. After each status update, I'll append one line to changelog.md:

2024-04-22 09:03 — thread_apr_22.txt: added API dependency blocker, updated data sync and frontend sections, recorded mock layer decision

The changelog grows over time. The status doc stays current. You get both: a snapshot of now and a record of how it got there.