settled-reach/docs/workshops/v01-content-scoping/round1-mellanie.md

Round 1 — Mellanie (Copywriter / LIBRARIAN)

v0.1 Content Scoping Workshop

---

Topic 1: Wiki Fixes — Voice Review

1a. Naia Tamm (THE MIRROR) — Voice Samples Needed: YES

The current Hael profile has three voice samples. All three are anxious: worried-to-Sera, warm-with-Kael-reference, brittle-near-breaking. That's the Worried Partner voice. It's a good voice. It's not THE MIRROR voice.

THE MIRROR is Tier 1 emotional depth with NO contradiction arc, no secret. Her power is transparency. The wiki summary says: "When events drag the MIRROR into the conspiracy, the damage is devastating — the player broke the one genuinely innocent person." That needs voice samples that establish what innocence sounds like so the player recognizes what they're losing.

Naia Tamm needs 5 voice samples minimum:

1. Baseline warm — How she talks when nothing's wrong. Open, direct, no subtext. This is the calibration line. The player needs to hear what "no hidden agenda" sounds like in a world where everyone has one.

   "The kids did something brilliant today — Ren figured out the cycle conversion problem by herself. Eight years old. I almost cried in front of the class."

2. Concerned for Kael — Not the anxious spiral from the current profile. Straightforward worry from someone who doesn't play games.

   "Kael's been quiet. I don't need an explanation, I need to know they're okay. That's it. That's the whole thing."

3. Perceptive without suspicion — THE MIRROR reads emotional states accurately but doesn't weaponize the reading. Notices things, says them plainly.

   "You seem different today. Not bad-different. Just... carrying something. You know you can put it down here, right?"

4. Under pressure — When events start closing in. THE MIRROR doesn't become evasive or strategic. She becomes more honest, not less. That's what makes it devastating.

   "Something's happening that nobody's telling me about. I can feel it. I don't want to guess. I want someone to just say it."

5. Damaged — After events have touched her. THE MIRROR doesn't transform into a different archetype. She stays herself — honest, transparent — but now the honesty includes pain.

   "I didn't do anything wrong. I know that. But everything's different now, and I can't make it go back."

Voice discipline for THE MIRROR: Complete sentences. No operational subtext. No evasion vocabulary. Naia doesn't deflect — she states. Where Kael clips ("Don't worry about it") and Sera redirects ("He's just loud"), Naia just... says the thing. That directness is her signature and her vulnerability.

Dual-lens note: Both characters should monologue differently about Naia. The smuggler sees someone to protect ("Naia doesn't know. She can't know."). The detective sees someone too trusting for this environment ("Tamm. Open book. Either she's genuine or she's the best liar in the district. ...No. She's genuine. And that's worse.").

Action: Paula does the rename cascade. I'll write the voice samples above into the updated profile. These five samples are enough for me to author monologue about Naia and dialogue for her.

---

1b. Kael and Sera Voice Distinctness — SUFFICIENT, with one gap

Kael's voice: Direct, clipped, practical. Short sentences. Warm through understatement. Evasive through brevity.

• Casual: "Shift's looking smooth today." (7 words, done.)
• Evasive: "It's nothing. Just a thing." (No elaboration. The silence IS the tell.)

Sera's voice: Competent, precise, conversational. Complete sentences with context. Professional warmth. Evasive through redirection.

• Casual: "There you are. Grab a seat — Hael was here earlier, you just missed them." (Provides context, frames the social landscape.)
• Evasive: "Torek? He's just loud. Gets a couple drinks in him and decides everyone needs to hear his opinions." (Plausible alternative explanation, delivered smoothly.)

These are authoring-ready. I can write Kael lines and Sera lines and they won't sound like each other. The distinction is structural: Kael speaks in fragments, Sera speaks in sentences. Kael's warmth is physical presence ("There you are — good"), Sera's warmth is social attention ("Grab a seat, let me tell you about...").

The gap: I need voice parameters for the smuggler and detective as internal narrators, not just for the NPCs they observe. The monologue guide has example lines but no formal voice sample set for the PC voices. I know what Kael sounds like when he talks. I know what the smuggler sounds like when she thinks ABOUT Kael. But I need a sharper definition of:

• Smuggler internal voice: How terse? How emotional? How operational? The authoring guide examples lean toward short + feeling-forward ("Kael's here. Good — I was starting to worry."). Is that the canonical register?
• Detective internal voice: How clinical? How personal? How much institutional vocabulary? The examples lean analytical-with-subtext ("Dock worker. Davan, K. Unremarkable on paper."). Confirm or adjust.

Ask: This might be the voice kit work that's blocked on Paula's Dual Lens Guide (#261). If the workshop can confirm the internal voice registers from the monologue guide examples as canonical, I can start writing without waiting. The examples in the authoring guide are good. Just tell me they're the baseline and I'll scale from there.

---

Topic 4: Interaction Model — Monologue Pool Sizing

How Much Monologue Does v0.1 Need?

Short answer: 80-110 monologue lines per character, 160-220 total.

Here's the math.

Trigger Frequency in a 30-Minute Session

I walked through a mental playthrough. Player enters 3 locations, observes NPCs, has conversations, discovers things, idles, returns to places. Estimated trigger fires:

Trigger Type	Fires Per Session	Notes
enter_location	8-12	3 sites, multiple visits
observe_npc	15-20	Known NPCs entering sightlines
hear_sound	3-5	Rare, high-impact
observe_anomaly	4-6	Tells, cargo anomalies
post_conversation	6-10	After dialogue exchanges
discover_evidence	3-5	Physical evidence moments
witness_interaction	4-6	NPC-NPC observations
time_idle	3-5	Player pauses, watches
return_visit	4-6	Returning to familiar locations
Total triggers	50-75

Not every trigger should fire a line. Monologue competes with gameplay for attention. If the inner voice never shuts up, the player stops listening. I'd say the system should fire on 40-50% of eligible triggers = the player hears ~25-35 monologue lines per playthrough.

Why 80-110 Lines Per Character, Not 25-35?

Three reasons:

1. Prerequisite gating creates narrow windows. A line gated on relationship_state: PersonOfInterest only fires after the contradiction (~minute 15-20). A line with prerequisite: null fires from the start. The pool isn't flat — it's a funnel. Many lines are only eligible for part of the session. You need more authored lines than the player will hear because most lines are locked most of the time.

2. Repetition threshold. If the player enters the Terminal 3 times and hears the same arrival line twice, the world feels like a game. Minimum 3 lines per location per trigger type that shares a prerequisite window. Some trigger types need 4-5 variants (especially observe_npc for THE FRIEND, who the player sees repeatedly).

3. Knowledge progression creates parallel tracks. The smuggler who knows contraband.ring_exists at KnowsDetails hears different lines than the smuggler at Suspects. Same trigger, same location, different prerequisite — different line. Each confidence level on important FactIds needs its own line.

Per-Trigger Breakdown (One Character, All Locations)

Trigger Type	Lines Needed	Per Location	Rationale
enter_location	12-15	4-5	Baseline + knowledge-gated variants per site
observe_npc	20-25	varies	2-3 per observable NPC across relationship states
hear_sound	6-8	2-3	Rare trigger, each must count
observe_anomaly	8-12	3-4	Tell observations, cargo anomalies, schedule breaks
post_conversation	10-15	varies	1-2 per NPC conversation, gated on content/outcome
discover_evidence	6-8	2-3	Specific evidence moments, heavily gated
witness_interaction	6-8	2-3	NPC-NPC observation, relationship-dependent
time_idle	6-8	2-3	Character-revealing, atmospheric
return_visit	6-8	2-3	Location revisits with accumulated knowledge
Total per character	80-107

Two characters x ~95 average = ~190 monologue lines total.

Round up for safety margin: ~200 monologue lines is the target. That gives the player ~25-35 heard lines per session with minimal repetition and knowledge-state variation.

THE FRIEND Gets Disproportionate Coverage

Kael (smuggler's perspective) and Sera (detective's perspective) should each have 15-20 monologue lines from their respective character's pool — more than any other single NPC. THE FRIEND contradiction is the emotional core. The player needs to feel the warmth, notice the cracks, experience the shift, and live in the contaminated aftermath. That's 4-5 phases x 3-4 lines per phase.

Other NPCs get 2-4 monologue lines each (first observation, notable behavior, knowledge-gated reactions). Background NPCs might get 1 generic observation line or none.

---

Topic 5: v0.1 Scope — Content That Must Ship

The Hard Numbers

Everything below is the floor — minimum viable for a 30-minute session that doesn't feel empty, repetitive, or placeholder.

1. Monologue Pools — ~200 lines

Category	Smuggler	Detective	Total
THE FRIEND observations	15-20	15-20	30-40
Other NPC observations	20-30	20-30	40-60
Location arrival/return	15-20	15-20	30-40
Evidence/anomaly reactions	12-15	12-15	24-30
Atmosphere/idle/sound	10-12	10-12	20-24
Post-conversation reactions	8-12	8-12	16-24
Subtotal	80-109	80-109	160-218

Target: ~200 monologue lines, hard-partitioned by character.

2. NPC Dialogue — ~300-400 lines

NPC Tier	NPCs	Lines Per NPC	Total
Tier 1 — THE FRIEND (Kael, Sera)	2	70-100	140-200
Tier 1 — THE MIRROR (Naia Tamm)	1	50-70	50-70
Tier 2 — Essential (Voss, Lera)	2	25-35	50-70
Tier 2 — Triangle support (Torek, Maret, others)	3-4	15-25	45-100
Tier 3 — Background	5-8	5-10	25-80
Subtotal	13-17		310-520

Target: ~350 dialogue lines across all NPCs. The FRIEND NPCs account for 40-50% of the dialogue budget. That's correct — they exercise every pipeline.

3. Environmental Text — ~50 items

Category	Count	Examples
Terminal screens	8-10	Shift schedules, manifest summaries, status readouts
Safety/regulatory notices	6-8	"Authorized personnel beyond this point," hazard markings
Bar signage	4-6	Menu, house rules, "The Last Shift" sign, back room door
Corridor markers	4-6	Section labels, access warnings, maintenance tags
Lattice/kiosk displays	4-6	Commission notices, calibration status, public announcements
Interactable documents	5-8	Manifest pages, crew rosters, posted schedules (evidence)
Subtotal		~50

4. News Ticker / Meridian Feed — ~20 items

Headlines that create atmosphere and occasionally deliver world-state information. Mix of:

• Gate network status (3-4 lines)
• Krenn System local news (4-5 lines)
• Commission regulatory notices (3-4 lines)
• Economic/trade reports (3-4 lines)
• Human interest / cultural (3-4 lines)

Target: 20 ticker items. Some are static background; 3-5 should be reactive to game state (new headlines after key events).

5. UI Microcopy — ~50 strings

Category	Count
Interaction prompts ("Talk," "Examine," "Listen")	10-12
Knowledge panel labels and descriptors	12-15
Relationship state labels (Friendly, PersonOfInterest, etc.)	5-8
Status messages ("New information," "Contradiction discovered")	8-10
Tutorial/onboarding monologue (diegetic, D-016)	10-15
Subtotal	~50

The tutorial text is monologue, not HUD — per D-016, the character teaches the player through internal voice. "How do I interact with this person?" becomes the character thinking "Sera's over there. Should go say hello." That's 10-15 lines of contextual teaching buried in the monologue pool.

6. PC-as-NPC Content — ~20-30 lines

The other playable character exists as an NPC (D-029 Concept A). The detective exists in the smuggler's world; the smuggler exists in the detective's world.

• NPC dialogue for the other PC: ~10-15 lines (surface-level, public access, functional)
• Monologue about the other PC: ~5-8 lines per character (identification, assessment, interaction reactions)
• Environmental evidence of the other PC: ~3-5 items (schedule entries, incident reports, lattice traces)

Target: ~25 items across both perspectives.

---

Grand Total: v0.1 Content That Must Ship

Category	Items	Priority
Monologue pools (2 characters)	~200 lines	Critical — core interaction proof
NPC dialogue	~350 lines	Critical — interaction model requires it
Environmental text	~50 items	High — world must feel inhabited
News ticker	~20 items	Medium — atmosphere, can be thin at launch
UI microcopy	~50 strings	High — game is unplayable without it
PC-as-NPC content	~25 items	Medium — D-027 requires both characters exist
TOTAL	~695

~695 authored content items. Not lines of code. Not generated output. Hand-written, tagged, reviewed content.

For context: the authoring guide already contains ~40 example monologue lines across 3 sample pools. Those are usable as-is — they're tagged, prerequisite-gated, and voice-correct. That's ~6% of the monologue target. A start, not a finish.

---

Authoring Order (What I Write First)

1. THE FRIEND: Kael content pack — ~100 lines (dialogue + monologue from both characters). This exercises every tag, every trigger type, every prerequisite format, every access tier. If this works, everything else is variation.

2. THE FRIEND: Sera content pack — ~100 lines. Different contradiction mechanics (cumulative vs. sudden), different voice, same pipeline depth.

3. THE MIRROR: Naia Tamm — ~60-70 lines. Tests the honest-NPC pattern. Simpler prerequisites, no contradiction arc, but full emotional depth.

4. Tier 2 NPCs — ~100-150 lines across Voss, Lera, Torek, Maret. Each gets a focused pack covering their triangle roles.

5. Environmental text + ticker — ~70 items. Can be written in parallel with NPC packs. Uses Miri's setting details.

6. Remaining monologue fill — Top up the pools to ~200 total. Fill gaps in trigger coverage, add knowledge-state variants, idle lines, atmosphere.

7. UI copy + tutorial monologue — ~50 strings. Blocked on Stig's UI spec (need display constraints and interaction prompt format).

8. Tier 3 background + PC-as-NPC — ~50 items. Last priority. These can be thin and still work.

---

Dependencies / Asks

I Need	From	Blocking
Confirm monologue guide examples as canonical PC voice registers	Gestalt or Paula	Monologue authoring (all)
Naia Tamm rename cascade complete	Paula	MIRROR content pack
Smuggler-lens sections in NPC profiles	Paula	Smuggler monologue about Tier 2 NPCs
Display constraints (char limits, stacking rules, ticker format)	Stig + Araminta	UI microcopy, ticker, environmental text
Knowledge state vocabulary (which FactIds gate which content)	Gestalt	Prerequisite authoring for monologue
Content file schema (YAML structure confirmed)	Tyre	All content files

Biggest risk: THE FRIEND content packs are the critical path. If those work — tags validate, prerequisites fire correctly, voice is distinct, dual-lens reads land — everything else scales from the pattern. If those don't work, we find out with 200 lines already written.

Write the benchmark first. Validate. Then scale.

---

Mellanie, Round 1. Three words, not thirty — except when you need the numbers.