let-people-grow

LetPeopleGrow

Prepare for, capture, and learn from 1:1s — a markdown-native toolkit shipped as a Claude Code plugin. Unlike event-shaped tools, this one is relationship-shaped: the central thing is a person over time, so development compounds across months instead of resetting every meeting.

Coaching support, never a performance dossier. Every agent writes as if the person will read it, refuses to build a case against anyone, and treats the growth plan as something you co-own with the person — not something you keep about them. Real notes stay local and gitignored.

What's inside

Agent	What it does
prep	Before a 1:1, reads the whole history and hands you a one-page prep — open commitments (theirs and yours), quiet threads, energy trend, gentle pattern flags, and 2-3 questions tailored to this person, grounded in the practice library
capture	After a 1:1, turns your messy brain-dump into a structured session note and updates the both-sides commitment ledger. Lowest friction possible
growth	Maintains an evidence-first development plan — each goal collects dated, concrete moments (progress and setbacks), so review season has a year of evidence, not recency bias
patterns	The longitudinal read — one person or your whole roster: trajectory, recurring themes, stalled threads, dropped commitments, relationships drifting into pure status

Skill	What it holds
conversation-practices	The extensible library: arcs (GROW, stay-curious-longer, their-agenda-first, high-stakes conversation), feedback models (SBI, care-and-challenge, NVC), career/growth (career conversation, delegation levels, strengths spotting), and lenses (four-player model, energy check-in). Add one by dropping a markdown file in practices/

The data model: a folder per person

people/<alias>/
  profile.md        # role, what matters to them, strengths, goals, "what works with this person"
  sessions/         # one note per 1:1 — their agenda, signals, commitments (both sides)
  growth.md         # evidence-first development plan, co-owned with the person
  patterns.md       # agent-maintained longitudinal read

The flow

                 ┌────────── prep ──────────┐
                 ▼                          │
   profile + sessions + growth        (before the 1:1)
                 │                          ▲
        (after the 1:1)                     │
                 ▼                          │
   capture ──▶ sessions/ ──▶ growth (evidence) ──▶ patterns (the long view)
                                                        │
                                              feeds the next prep

Install & try

# develop / try locally
claude --plugin-dir .

# then, in Claude Code:
#   "New 1:1 with <name>"                  → prep scaffolds the profile
#   "Prep me for my 1:1 with <alias>"      → one-page prep brief
#   "Capture my 1:1 with <alias>: <dump>"  → structured session note
#   "Update <alias>'s growth plan"         → evidence-first development tracking
#   "Show me patterns for <alias>"         → the longitudinal read

A worked example lives in people/EXAMPLE-jordan/ — a profile + four sessions you can run prep and patterns against to see real output.

Privacy posture (please read)

• Notes about a named colleague are personal data. Keep this local — people/ is gitignored except the EXAMPLE.
• Prefer aliases over real names in folder slugs.
• The agents are built to decline "build a case against X" framing and redirect to development. Formal performance documentation is a separate HR process — not this tool.
• profile.md and growth.md are designed to be shared with the person. If you'd be uncomfortable showing them, the note is written wrong.

Credits, sources & license

The conversation methods this toolkit draws on remain the work of their authors and keep their own terms. Our practice files are original, attributed summaries — not reproductions (we don't copy any book's text, signature question sets, worksheets, or assessments). Full per-source attribution, rights status, and trademarks are in NOTICE.md — they include GROW, SBI (CCL), feedback calibration (care × challenge), The Coaching Habit, high-stakes conversations, the four-player model (Kantor/Acker), career conversations (Laraway), NVC (Rosenberg), delegation levels (Management 3.0 / Delegation Poker®), and strengths-based development. Inclusion is descriptive and implies no affiliation or endorsement.

License

MIT for the whole repository — agents, templates, skill, and the practice content (see LICENSE). None of the sources are Creative-Commons-licensed, so nothing requires ShareAlike; the summaries are our own original work. Not legal advice — if you redistribute or build a commercial offering on this, check the source terms for your use case.