Sat 29 Aug · 19:25 IST

Day 10 / foundation

Senior PM (SPM) at big tech. Product management, consumer + AI. · Full System (solo-first, FOUNDATION mode). · Senior (9 yrs PM/analytics; PM2 at Zynga, 4 yrs).

The one thing now

pull-based metrics retrieval (name one existing story + 10s opening line) + self-rate 1-5. Issued 2026-08-28.

pick story for "moved a metric" prompt, name it, give one-line opening, self-rate 1-5. On reply: log story + score. If no reply, easy stop: prompt-swap next.

Morning floor · 8 min · strong window

repeat, graduating difficulty when 4-5 on a category 2w in a row

Evening split · light · optional

01 Standing

Storybank

0/8

gate to advanced drills

Cold baseline

none

2-weekly unassisted mock

Sessions

state 2026-08-25

Automation

5 jobs

delivery failure — see system

02 Gaps

Storybank 0/8 — the gate to advanced drills

No cold baseline yet — the 2-weekly unassisted mock hasn't run

Coaching state 4d old — answer the pending micro-step to close the loop

Automation delivery failure — see system

03 Refresher — keep warm

Answer shape under pressure

Every behavioral answer lands in one shape: context in one breath → decision → your specific move → measured outcome. If you can't name the number, you don't have the story yet. Land the number before the nuance.

Pacing — your known leak

History shows weak pacing and early "not-so-much" signals from interviewers. Rule: 10 seconds of silence to think is fine. Open with the headline, then support it. Never narrate your way into a point.

The PM2 → SPM story

Interviewers will probe platform-vs-feature depth and org-level influence. Every story should end with scope that grew: what you influenced beyond your lane, and who followed. Close the narrative gap before it's poked.

Impostor spiral antidote

Rust is not decline. The system's bet: retrieval beats review. A 5-minute pull drill on a bad day counts more than a perfect study block planned for tomorrow. Log the attempt, not the shame.

The rubric you're scored on

SPM bar = structured thinking · product judgment · execution credibility · influence · communication. Self-rate each drill 1–5 against one of these dimensions, not all five at once.

04 The process — you are here

TRUNK — solo, daily (always available)

- Morning (primary, strong window): 8-minute daily floor via `interview-coach`. One micro-drill. One visible number score. `practice` (1 round, ~5 min) OR 1 rapid `stories` retrieve OR 1 mock question. - Evening (lighter split): story retrieval drill or one `practice` round + `progress` check. Optional, 40% intensity of morning.

REPPS — real-interview simulation (2 sources)

1. Lenny's guests (`lenny-guest-list` skill, installed earlier this thread): 1-2 company-appropriate tough interviews/week. Good when you want an interviewer who won't flatter you and matches the role (AI PM, growth, strategy). 2. Akanksha — reactivates when she's back. No fixed cadence until then.

EVENT-DRIVEN — only when a real interview is scheduled

`research [company]` → `prep [company]` → `concerns` → `hype` → `questions` → one `mock`. Deadlines override procrastination, so this mode works well. We don't run it empty.

AUTOMATION — fixed to match the trunk

- 8:00 AM daily: valid priming question + micro-drill from `interview-coach` state (NOT a dead rotation plan). - 9:30 AM: re-enabled, pointing at interview-coach progress (was paused for a reason). - Sunday 21:00: real weekly review — fix the one that errored Aug 16.

System map — trunk → gate → reps open full map ↗

05 Operating principles

**Attempt-first, always.** Answer before you look anything up. Retrieval is the workout; review is just stretching.

**One micro-step at a time.** The coach states a single next action. No menus, no choice paralysis.

**Every turn ends in a number.** Score 1-5, storybank count, streak — visible progress, no vibes.

**Evidence, not shame.** Errors go in the ledger with cause + handling. Recurrence is data, not failure.

**Morning is the strong window.** Hard drills before noon; evening is intentionally lighter (~40%).

**Easy stop allowed.** If a drill doesn't land, say stop and the prompt swaps. Friction is a bug, not a virtue.

**Batch storybank to 8.** The gate to advanced drills. One story per drill is enough progress for a day.

06 System

Automation health

Job	Schedule	Health	Last run	Next	Channel
Daily Muesli PM Mock Sync	30 9 * * 1-5	paused	29 Jul 09:30	30 Jul 09:30	telegram:1098603668
Morning Prep Floor (8 AM)	0 8 * * 1-5	healthy	28 Aug 08:00	31 Aug 08:00	kapso
Sunday Solo-First Weekly Review	0 21 * * 0	DELIVERY FAIL live adapter send failed: Kapso API HTTP 422: {"error": "Cannot send non-template messages outside the 24-hour	23 Aug 22:07	30 Aug 21:00	kapso
Evening Light Split	0 20 * * 1,3,5	error	28 Aug 20:25	31 Aug 20:00	kapso
Prep Delivery Watchdog	every 180m	healthy	29 Aug 18:09	29 Aug 21:09	telegram:1098603668

Session log — 6

Date	What ran	Outcome
2026-08-28	morning floor (swap to pull-based retrieval)	Two retros answered. Swapped to pull-based drill per Coaching Note to cut friction: pick existing story for metrics prompt + 10s opening, self-rate 1-5. Awaiting reply; log story + score on response.
2026-08-27	morning floor (story surfacing)	Re-issued metrics retrospective prompt (storybank still <8). Awaiting reply; score + Error Ledger row on response.
2026-08-25	morning floor (story surfacing)	Sent retrospective prompt to surface a metrics story (moved a number). Attempt-first, self-rate 1-5. Awaiting reply; score + Error Ledger row when it lands.
2026-08-24	evening split (story retrieval drill)	Prompted solo retrieval drill: pick right story for a stakeholder-conflict prompt + 10s opening line, attempt-first, self-rate 1-5. Awaiting reply; score and Error Ledger on response. Next micro-step set to same drill with metrics prompt.
2026-08-21	morning floor	Acknowledged cron job receipt. Foundation cadence continues.
2026-08-20	system setup	Trunk created: solo-first FOUNDATION system. Akanksha on hold. Morning floor + evening split + Sunday review live.

Verified working methods

Method	Works when	Evidence
None verified yet — methods graduate here once evidence lands.

Coaching notes

ADHD: needs micro-steps (<5 min), no-choice greetings, visible number scores, no shame. See references/adhd-overrides.md.

Strongest focus window: morning. Evening is lighter.

Gap: hasn't had a real mock in months (Akanksha traveling). Rebuild story retrieval first.

Signals possible stall on story surfacing drills (08-25 prompt unanswered). If second retro prompt gets no reply, swap drill type to pull-based retrieval (name the story for a metrics prompt) to reduce friction.

Error ledger — 0

Date	Error/Gap	Evidence	Cause	Handling	Recurs
Clean — no logged errors yet. Evidence, not shame: log the first one when it happens.

Interviewer's lens

They'll like: $20M portfolio ownership, $1M+ opportunity discovery, multi-year revenue turnarounds, 0→1 features, A/B testing, stakeholder mgmt.

They'll probe: depth of platform vs feature work; org-level influence; technical depth (counter: AI/LLM learning is a strength).

Narrative to close: build the 'PM2→SPM' story; platform-thinking proof points.

Your read: impostor spiral; gap since last real sim; weak pacing/early NSM historically. · anxiety: post-gap / impostor-lean. · history: active in past but system dormant 6+ weeks. Experienced, rusty.

Baseline rule: one timed solo mock, self-scored on SPM rubric, stored at `baselines/baseline-YYYYMMDD.md`

Sources: coaching_state.md · PREP-SYSTEM.md · REFRESHERS.md · cron state — updated 29 Aug 2026 19:25 IST
Republish: bash publish_dashboard.sh → prep.aravind.app