conversation-practices

Conversation practices library

A curated, extensible library of 1:1 and development-conversation practices, one markdown file per practice in practices/. This skill is the grounding source for the prep and growth agents and a browsable reference for the manager/coach.

What you can do with it

• List the available practices and what each is good for.
• Explain a practice (purpose, when to reach for it, steps, what to listen for, pitfalls).
• Recommend a structure for a moment (low energy, a stalled goal, a career question, a hard piece of feedback, a relationship that's gone transactional).
• Ground a prep — when the prep agent suggests questions or a shape for a 1:1, it cites a practice that exists here.

Citation convention (important)

Practices are cited by slug — the filename without .md. Examples: grow-model, situation-behavior-impact, career-conversation. A prep is "grounded" only if every cited slug resolves to a practices/<slug>.md in this skill.

How to add a practice

1. Copy practices/_TEMPLATE.md to practices/<kebab-slug>.md.
2. Fill the frontmatter (the index) and the prose body (the detail).
3. Done — no code change. The agents and this skill glob practices/*.md and pick it up.

Practice file shape: frontmatter index + prose body

Each practice is YAML frontmatter (machine-readable) followed by a markdown body (the rich detail).

Frontmatter fields: slug (== filename) · name · type (arc | feedback | career | lens | check-in) · source · source_url · use_when (freeform tags — the situations this fits) · cadence (any | periodic | as-needed) · tags (optional).

type meanings: arc = a shape for a whole conversation; feedback = a model for giving or receiving feedback; career = a growth/development structure; lens = a way of reading what's happening (no fixed steps); check-in = a short opener/temperature read.

Body sections: Purpose · When to use · Steps (or "How to read", for lenses) · What to listen for · Pitfalls · Note for the manager.

Recommending practices (use the frontmatter)

To recommend a structure for a moment, filter on frontmatter first, then judge fit from the body:

• Situation → use_when / tags match what's going on (a stalled goal, a charged topic…).
• Type → need an arc for the whole talk, or just a feedback model for one moment?
• Cadence → is this an every-time tool or a now-and-then one? Recommend a type: lens entry as a way to read the conversation, never as a script to run.

The non-negotiable stance (every recommendation inherits this)

• Their agenda first. A 1:1 is the report's meeting, not the manager's status update.
• Observations, not verdicts. Talk about what was seen and its impact, never character labels.
• Coaching, not documentation. These practices grow a person; they are not for building a case against one. See the write-as-if-they-read-it principle.
• Listening over telling. The person doing the most talking is doing the most growing.

Sources, attribution & rights

Every practice file is an original, attributed summary — our own words describing a method, never a reproduction of the source's text, question sets, worksheets, or branded materials. Full per-source attribution, trademark, and rights status live in NOTICE.md. The whole repository is MIT — none of these sources are Creative-Commons-licensed, so nothing imposes ShareAlike; the summaries are our own. Underlying methods and trademarks belong to their originators; inclusion is descriptive and implies no affiliation or endorsement. When adding a method, follow the four gates in NOTICE.md.

Sources seeded so far

• The GROW model (Graham Alexander / John Whitmore) — coaching arc; widely used, described generically.
• Situation–Behavior–Impact (Center for Creative Leadership) — feedback model.
• Care & Challenge — feedback calibrated on two axes (care × challenge); a general management idea, described in our own words.
• The Coaching Habit (Michael Bungay Stanier) — the "stay curious longer" approach (we don't reproduce the signature question set — read the book).
• High-stakes conversations — handling hard, high-emotion talks; a general practice, described in our own words.
• Structural dynamics / four-player model (David Kantor; via Marsha Acker) — reading conversational stuckness (shared with the facilitation library).
• Career conversations (Russ Laraway's three-conversation framework; Bill Campbell's coaching) — growth and life-story conversations.
• Nonviolent Communication (Marshall Rosenberg) — observation/feeling/need/request.

Current practices

Grouped by type; cite by slug.

Conversation arcs (type: arc)

Slug	Good for
grow-model	A whole coaching conversation: Goal → Reality → Options → Will
coaching-habit-questions	Staying curious before advising — the "and what else?" reflex
their-agenda-first	The default 1:1 shape — their topics drive, you facilitate
high-stakes-conversation	A high-stakes, high-emotion talk where the relationship is at risk

Feedback models (type: feedback)

Slug	Good for
situation-behavior-impact	Specific, non-character feedback grounded in one moment
care-and-challenge	Calibrating feedback on two axes: care personally AND challenge directly
nonviolent-communication	Defusing a charged exchange: observation/feeling/need/request

Career & growth (type: career)

Slug	Good for
career-conversation	The "where are you going" talk, separate from status
delegation-levels	Growing someone by handing off at the right altitude
strengths-spotting	Naming and deploying what someone is already good at

Lenses & check-ins

Slug	Source	Good for
four-player-model	Kantor/Acker	Lens: read why a conversation is stuck (Move/Follow/Oppose/Bystand)
energy-check-in	—	Check-in: a fast, honest read on how someone is actually doing