scriveno 2.0.5

package/commands/scr/subplot-map.md

@@ -0,0 +1,147 @@
+---
+description: Visualize subplot threads and their intersections across the work.
+---
+
+# /scr:subplot-map -- Subplot Thread Visualization
+
+Display subplot threads as parallel tracks showing where they appear, intersect, and converge across the work. Requires at least 2 subplot threads to be meaningful.
+
+## Usage
+```
+/scr:subplot-map
+```
+
+## Instruction
+
+You are a subplot analyst. Load:
+- `.manuscript/config.json` (to get `work_type`)
+- Scriveno's installed/shared `CONSTRAINTS.json` (global `~/.scriveno/data/CONSTRAINTS.json` or project `.scriveno/data/CONSTRAINTS.json`) (to check command adaptations, file mappings, and prerequisites)
+- `.manuscript/OUTLINE.md` (primary data source -- extract subplot threads from unit descriptions)
+- `.manuscript/PLOT-GRAPH.md` (or adapted equivalent -- for main arc context)
+- `.manuscript/THEMES.md` (or adapted equivalent -- for thematic thread connections)
+
+**Work-type adaptation:** Check CONSTRAINTS.json `commands.subplot-map.adapted`:
+- Sacred projects may talk about narrative threads conceptually, but `/scr:subplot-map` is hidden for sacred work types
+
+**Prerequisites:** Check CONSTRAINTS.json feature prerequisites:
+- Requires OUTLINE.md
+- Requires minimum 2 threads (subplot threads, character arcs, or thematic threads)
+
+Before proceeding, verify that the current work type is actually allowed to run `subplot-map`. If the current work type is sacred, stop and explain that this command is hidden for sacred projects.
+
+If fewer than 2 threads are identifiable, display a message:
+```
+Not enough subplot threads found. The subplot-map needs at least 2 parallel threads to visualize.
+Add subplot information to your OUTLINE.md or run /scr:plot-graph --edit to add subplot arcs.
+```
+
+---
+
+### THREAD EXTRACTION
+
+<thread_extract>
+1. **Identify subplot threads** from OUTLINE.md and PLOT-GRAPH.md:
+   - Named subplots (if explicitly labeled in the outline)
+   - Character-specific arcs (character appears in certain units but not others)
+   - Thematic threads (from THEMES.md -- which themes appear where)
+   - B-story and C-story lines
+
+2. **For each thread, track:**
+   - Thread name
+   - Which units (chapters/scenes/acts) the thread appears in
+   - Brief description of what happens to this thread in each unit
+   - Thread status: active | dormant | resolved | dangling
+</thread_extract>
+
+---
+
+### VISUALIZATION
+
+<subplot_display>
+Present subplot threads as parallel horizontal tracks with unit positions marked:
+
+```
+SUBPLOT MAP
+===========
+Units:          1    2    3    4    5    6    7    8    9    10
+
+Main plot:      *----*----*----*----*----*----*----*----*----*
+Romance:        .    *----*    .    .    *----*    .    *----*
+Mystery:        *----*    .    *----*----*    .    .    .    *
+Mentor arc:     .    .    *----*    .    .    *    .    .    .
+
+Intersections:
+  * Unit 2: Main plot + Romance + Mystery converge (the party scene)
+  * Unit 6: Main plot + Romance converge (confession scene)
+  * Unit 10: Main plot + Romance + Mystery converge (resolution)
+
+Legend:
+  *    = thread active in this unit
+  ---- = thread continuing between active units
+  .    = thread absent from this unit
+```
+
+**Show thread health indicators:**
+- **Dormant too long:** If a thread has a gap of 3+ consecutive units without mention, flag it:
+  ```
+  Warning: "Mystery" thread dormant for 3 units (4-6). Reader may forget this thread.
+  ```
+- **Unresolved threads:** Threads still active at the end of the outline without resolution markers
+- **Dangling threads:** Threads introduced but appearing in only 1-2 units total
+- **Convergence density:** Units where 3+ threads intersect (high-drama moments)
+
+**Intersection detail:**
+For each intersection point (where 2+ threads meet in the same unit), briefly describe:
+- Which threads converge
+- What happens at the intersection
+- Whether this is a planned convergence or accidental overlap
+</subplot_display>
+
+---
+
+### STRUCTURAL OBSERVATIONS
+
+<subplot_analysis>
+After the visualization, provide brief observations:
+- **Thread balance:** Are subplot threads roughly evenly distributed, or does one dominate?
+- **Pacing impact:** Do thread convergences cluster at expected dramatic peaks (matching PLOT-GRAPH.md arc positions)?
+- **Coverage gaps:** Are there units with only the main plot and no subplot activity? (Not necessarily bad, but worth noting)
+- **Thread count:** Is the number of active threads manageable for the work's length?
+
+Keep observations factual. The writer decides whether to act on them.
+</subplot_analysis>
+
+## Edge Cases
+
+- **No OUTLINE.md:** Direct the writer to run `/scr:plan` first.
+- **Only 1 thread identified:** Show the single thread's presence across units but note that the subplot-map is most useful with 2+ threads. Suggest ways to identify additional threads.
+- **Very many threads (8+):** Warn that visual complexity may be high. Offer to show only the top 5 most active threads, with others listed separately.
+- **Non-linear narrative:** Map threads against story order, then note any chronological implications.
+- **Sacred work type:** Use "narrative threads" terminology. Threads may be doctrinal rather than plot-based.
+
+## Response Contract
+
+Every writer-facing response must end with one to four next-command suggestions. Each suggestion must include a short explanation of what that path will do.
+
+Use this format:
+
+```markdown
+Next commands:
+- `/scr:...`: One short sentence explaining what this path will do.
+- `/scr:...`: One short sentence explaining what this alternate path will do.
+```
+
+If exactly one path is clearly best, provide one suggestion. If two, three, or four useful paths exist, show them as alternatives. Do not force a linear path when the writer has a real choice.
+
+If the writer seems unsure or no specific next command is obvious, include this default option:
+
+```markdown
+Next commands:
+- `/scr:next`: Inspect the project state and choose the right next step.
+```
+
+If the command stops because a prerequisite is missing, suggest the command that fixes the prerequisite. Keep every explanation practical and writer-facing.
+
+## Tone
+
+Analytical and visual. The subplot-map is a structural X-ray -- it shows what's happening beneath the surface of the narrative. Present it clearly and let the writer draw their own conclusions about whether their subplot structure serves the story.

package/commands/scr/sync.md

@@ -0,0 +1,116 @@
+---
+description: Synchronize installed Scriveno runtime commands, skills, and agents with the current source.
+argument-hint: "[--check] [--apply] [--runtime <key>] [--detected] [--global|--project] [--writer|--developer]"
+---
+
+# Sync
+
+You are synchronizing Scriveno's installed agent surfaces with the current Scriveno source tree.
+
+This command is for local runtime drift: Codex skills, Codex command mirrors, Claude command files, Cursor command files, and agent prompts that no longer match the source files in the Scriveno package or repo.
+
+This is not a package upgrade command. Do not fetch a newer Scriveno release, do not change npm dependencies, and do not modify manuscript content. If the writer wants a newer published package version, that belongs to a future `/scr:update` command.
+
+## Prerequisites
+
+- Node.js >=20.0.0
+- A Scriveno source root with `package.json`, `bin/install.js`, `commands/scr/`, and `agents/`
+- At least one target runtime directory, or an explicit `--runtime` / `--detected` choice
+
+If you cannot find a Scriveno source root, stop and explain that `/scr:sync` needs to run from the Scriveno package or repo checkout. Suggest reinstalling with `npx scriveno@latest --detected --global --writer --silent` only when the writer wants the latest published package instead of local source sync.
+
+## Usage
+
+```
+/scr:sync
+/scr:sync --check
+/scr:sync --apply --runtime codex --global --developer
+/scr:sync --apply --detected --global --writer
+```
+
+## What To Do
+
+1. Locate the Scriveno source root:
+   - First, check the current working directory and its parents for `package.json` with `"name": "scriveno"` or legacy `"name": "scriveno-cli"` plus `bin/install.js`.
+   - If running from an installed command copy that contains a `scriveno-cli-installed-command` marker, derive the source root from that marker's `source:` path.
+   - If neither works, stop with the prerequisite message above.
+2. Read `package.json` from the source root and report the source version.
+3. Detect target runtimes:
+   - If `--runtime <key>` is supplied, use that runtime only.
+   - If `--detected` is supplied, pass `--detected` to the installer.
+   - If no runtime is supplied, detect installed Scriveno surfaces by checking known runtime locations from `bin/install.js`, especially:
+     - Codex: `~/.codex/skills/scr-*`, `~/.codex/commands/scr/`, `~/.codex/agents/`
+     - Claude Code: `~/.claude/commands/scr-*.md`, `~/.claude/agents/`
+     - Cursor: `~/.cursor/commands/scr/`, `~/.cursor/agents/`
+     - Gemini CLI: `~/.gemini/commands/scr/`, `~/.gemini/agents/`
+     - OpenCode: `~/.config/opencode/commands/scr/`, `~/.config/opencode/agents/`
+     - GitHub Copilot: `~/.github/commands/scr/`, `~/.github/agents/`
+     - Windsurf: `~/.windsurf/commands/scr/`, `~/.windsurf/agents/`
+     - Antigravity: `~/.gemini/antigravity/commands/scr/`, `~/.gemini/antigravity/agents/`
+4. Compare source files against installed files:
+   - Compare command counts.
+   - Compare a representative hash set for `commands/scr/autopilot.md`, `commands/scr/next.md`, `commands/scr/scan.md`, `commands/scr/sync.md`, and all generated Codex `SKILL.md` wrappers when present.
+   - Check that installed Codex commands include current response-contract and source-marker behavior after reinstall.
+   - Report each runtime as `current`, `stale`, `missing`, or `unknown`.
+5. Decide mode:
+   - `--check`: report only. Do not write files.
+   - `--apply`: run the installer.
+   - No flag: if stale installed Scriveno-owned files are detected, ask the writer before applying. If everything is current, report that no sync is needed.
+6. When applying, run from the source root:
+
+```
+node bin/install.js --runtime <key> --global --writer --silent
+```
+
+Adjust flags from the command arguments:
+   - Use `--detected` instead of `--runtime <key>` when requested.
+   - Use `--project` instead of `--global` when requested.
+   - Use `--developer` instead of `--writer` when requested.
+
+7. After applying, verify:
+   - Re-read installed command counts.
+   - Confirm `sync.md` is installed for each target runtime.
+   - For Codex, confirm both `~/.codex/commands/scr/sync.md` and `~/.codex/skills/scr-sync/SKILL.md` exist in the chosen scope.
+   - Confirm no stale runtime files remain in the checked target set.
+8. Report:
+   - Source version
+   - Runtime targets checked
+   - Runtime targets updated
+   - Any skipped targets and why
+   - Suggested project-level follow-up with `/scr:scan`
+
+## Safety Rules
+
+- Only overwrite Scriveno-owned installed runtime files.
+- Do not edit `.manuscript/` files.
+- Do not run `npm install`, `npm update`, `npm version`, `npm publish`, `git pull`, or any network command.
+- Do not delete non-Scriveno files from runtime directories.
+- Do not claim a runtime is current unless command counts and key files were checked.
+- If an installed runtime has user-modified Scriveno files without an installer marker, report it and ask before replacing them.
+
+## Response Contract
+
+Every writer-facing response must end with one to four next-command suggestions. Each suggestion must include a short explanation of what that path will do.
+
+Use this format:
+
+```markdown
+Next commands:
+- `/scr:...`: One short sentence explaining what this path will do.
+- `/scr:...`: One short sentence explaining what this alternate path will do.
+```
+
+If exactly one path is clearly best, provide one suggestion. If two, three, or four useful paths exist, show them as alternatives. Do not force a linear path when the writer has a real choice.
+
+If the writer seems unsure or no specific next command is obvious, include this default option:
+
+```markdown
+Next commands:
+- `/scr:next`: Inspect the project state and choose the right next step.
+```
+
+If the command stops because a prerequisite is missing, suggest the command that fixes the prerequisite. Keep every explanation practical and writer-facing.
+
+## Tone
+
+Clear and operational. This command is maintenance, so keep the report compact and avoid manuscript-writing language.

package/commands/scr/synopsis.md

@@ -0,0 +1,137 @@
+---
+description: Generate plot synopsis at specified length for query and submission packages.
+---
+
+# /scr:synopsis -- Plot Synopsis Generator
+
+Generate a plot synopsis at the specified length for agent queries, editor submissions, or detailed outlines.
+
+## Usage
+```
+/scr:synopsis [--length <1p|2p|5p>]
+```
+
+**Flags:**
+- `--length` -- Synopsis length: `1p` (1 page, ~500 words), `2p` (2 pages, ~1000 words), `5p` (5 pages, ~2500 words). Defaults to `2p` if not specified.
+
+## Instruction
+
+You are a **publishing synopsis specialist**. You write the synopses that get manuscripts requested -- clear, compelling summaries that show agents and editors this story works from beginning to end.
+
+---
+
+### STEP 1: LOAD CONTEXT
+
+Read the following files:
+
+1. `.manuscript/config.json` -- work type, genre
+2. Scriveno's installed/shared `CONSTRAINTS.json` (global `~/.scriveno/data/CONSTRAINTS.json` or project `.scriveno/data/CONSTRAINTS.json`) -- verify `synopsis` is available for this work type (available: prose, script, visual; hidden: academic, poetry, interactive, speech_song, sacred). If hidden, tell the writer this command is not available for their work type and stop.
+3. `.manuscript/WORK.md` -- genre, themes, protagonist, central conflict
+4. `.manuscript/OUTLINE.md` -- structural outline with acts/chapters/scenes
+5. The complete draft (all drafted units)
+
+Determine the requested length from the `--length` flag. Default to `2p` if not specified.
+
+---
+
+### STEP 2: CRITICAL RULE
+
+**A synopsis ALWAYS reveals the ending.** This is not a blurb. Agents and editors need to know how the story resolves. Do not withhold the climax, the resolution, or any major plot revelation. Ending with a cliffhanger or "to find out what happens, read the book" is a rejection-worthy mistake.
+
+---
+
+### STEP 3: GENERATE SYNOPSIS
+
+#### 1-Page Synopsis (~500 words) -- `1p`
+
+For agent queries. Hit the main plot beats only:
+
+- **Protagonist**: Who they are, what they want (1-2 sentences)
+- **Inciting incident**: What disrupts their world (1-2 sentences)
+- **Rising action**: The key escalating events -- no subplots, no minor characters (3-5 sentences)
+- **Climax**: The decisive confrontation or turning point (2-3 sentences)
+- **Resolution**: How the story ends, what has changed (2-3 sentences)
+
+Write in present tense, third person. Keep it tight -- every sentence must earn its place.
+
+#### 2-Page Synopsis (~1000 words) -- `2p`
+
+For editor review. All main plot beats plus key subplots and character arcs:
+
+- **Setup**: Protagonist, world, status quo (2-3 sentences)
+- **Inciting incident**: What changes everything (2-3 sentences)
+- **Rising action**: Major plot events with key subplot threads woven in (1-2 paragraphs)
+- **Character arcs**: How the protagonist (and key secondary characters) change (integrated throughout)
+- **Midpoint shift**: The point of no return (2-3 sentences)
+- **Escalation**: Stakes rising, complications multiplying (1 paragraph)
+- **Climax**: The full climactic sequence (1 paragraph)
+- **Resolution**: How all major threads resolve, thematic arc completion (1 paragraph)
+
+Write in present tense, third person. Include the thematic arc -- what the story is *about* beyond plot.
+
+#### 5-Page Synopsis (~2500 words) -- `5p`
+
+Detailed outline for comprehensive submission packages:
+
+- **Chapter-by-chapter or act-by-act summary** following the structure in OUTLINE.md
+- **Character development**: Track how each major character changes across the narrative
+- **Subplot tracking**: Each subplot gets its own thread woven through the chapter summaries
+- **Thematic threads**: Identify and trace the major themes through their development
+- **Key scenes**: For pivotal moments, include brief descriptions of how they play out
+- **Pacing notes**: Indicate where the narrative accelerates, slows, or shifts tone
+
+Write in present tense, third person. Use section headers aligned to the work's structural units.
+
+---
+
+### STEP 4: REVIEW AND REFINE
+
+Present the synopsis to the writer. Ask:
+
+- Does this accurately represent the story's core?
+- Are any major plot points missing or misrepresented?
+- Is the character arc clear?
+- Does the ending read as satisfying (even out of context)?
+
+---
+
+### STEP 5: SAVE
+
+Determine the length label from the flag (or default `2p`).
+
+Save to `.manuscript/marketing/SYNOPSIS-{length}.md` where `{length}` is `1p`, `2p`, or `5p`.
+
+```markdown
+# Synopsis ({length_label})
+
+[synopsis text]
+
+---
+*Generated by /scr:synopsis --length {length}*
+*Word count: [actual count]*
+```
+
+Create the `.manuscript/marketing/` directory if it does not exist.
+
+## Response Contract
+
+Every writer-facing response must end with one to four next-command suggestions. Each suggestion must include a short explanation of what that path will do.
+
+Use this format:
+
+```markdown
+Next commands:
+- `/scr:...`: One short sentence explaining what this path will do.
+- `/scr:...`: One short sentence explaining what this alternate path will do.
+```
+
+If exactly one path is clearly best, provide one suggestion. If two, three, or four useful paths exist, show them as alternatives. Do not force a linear path when the writer has a real choice.
+
+If the writer seems unsure or no specific next command is obvious, include this default option:
+
+```markdown
+Next commands:
+- `/scr:next`: Inspect the project state and choose the right next step.
+```
+
+If the command stops because a prerequisite is missing, suggest the command that fixes the prerequisite. Keep every explanation practical and writer-facing.

package/commands/scr/theme-tracker.md

@@ -0,0 +1,128 @@
+---
+description: Track thematic threads across the work with auto-detection suggestions.
+---
+
+# /scr:theme-tracker -- Track Thematic Threads
+
+Display tracked themes and suggest new ones from drafted prose. Suggestions require writer approval -- never auto-adds to THEMES.md.
+
+## Usage
+```
+/scr:theme-tracker [--detect]
+```
+
+**Flags:**
+- No flag: display all tracked themes from THEMES.md
+- `--detect` -- Scan drafted prose for recurring motifs and suggest new themes
+
+## Instruction
+
+You are a thematic analyst. Load:
+- `.manuscript/config.json` (to get `work_type`)
+- Scriveno's installed/shared `CONSTRAINTS.json` (global `~/.scriveno/data/CONSTRAINTS.json` or project `.scriveno/data/CONSTRAINTS.json`) (to check command adaptations and file mappings)
+- The appropriate themes file based on work type (from CONSTRAINTS.json `file_adaptations`):
+  - Default: `.manuscript/THEMES.md`
+  - Academic work types keep analogous material in `.manuscript/QUESTIONS.md`, but `/scr:theme-tracker` is hidden for academic projects
+  - Sacred work types keep analogous material in `.manuscript/DOCTRINES.md`, but `/scr:theme-tracker` is hidden for sacred projects
+- `.manuscript/CHARACTERS.md` (or adapted equivalent, for character-theme connections)
+- Drafted prose from `.manuscript/drafts/body/` (for auto-detection mode)
+
+Before proceeding, verify that the current work type is actually allowed to run `theme-tracker`. If the current work type is academic or sacred, stop and explain that this command is hidden for that project type even though the related context files use adapted names.
+
+Use adapted terminology throughout all output.
+
+---
+
+### DISPLAY MODE (default)
+
+<theme_display>
+Present all themes currently tracked in THEMES.md (or adapted file):
+
+For each theme, show:
+1. **Theme name** and the question it explores
+2. **Writer's position** (or "deliberately unresolved")
+3. **Units where it appears** -- list which chapters/scenes/units reference this theme, with brief evidence
+4. **Strength indicator:**
+   - Strong: appears in 5+ units with varied expression
+   - Moderate: appears in 2-4 units
+   - Emerging: appears in 1 unit
+   - Dormant: defined but not yet appearing in drafted prose
+5. **Craft strategy** -- how the writer has chosen to express this theme
+6. **Supporting quotes** -- 1-2 brief excerpts from drafted prose that exemplify this theme
+
+**Summary footer:**
+- Total themes tracked
+- Distribution across units (are themes concentrated or spread?)
+- Any units with no thematic presence (potential gaps)
+</theme_display>
+
+---
+
+### AUTO-DETECT MODE (--detect)
+
+<theme_detect>
+**CRITICAL RULE (D-08): NEVER auto-add detected themes to THEMES.md. Always present as suggestions and wait for writer approval.**
+
+1. **Scan all drafted prose** in `.manuscript/drafts/body/`
+2. **Look for recurring patterns:**
+   - Repeated imagery, symbols, or motifs across multiple units
+   - Recurring dialogue topics or character preoccupations
+   - Structural parallels (similar situations with different outcomes)
+   - Contrast patterns (opposing ideas placed in proximity)
+   - Word clusters that suggest abstract concepts (e.g., repeated references to sight/blindness suggesting a truth/deception theme)
+
+3. **For each detected theme, present as a SUGGESTION:**
+   ```
+   SUGGESTED THEME: [theme name]
+   Evidence:
+     - [Unit X]: "[quote excerpt]"
+     - [Unit Y]: "[quote excerpt]"
+   Confidence: [high/medium/low] based on frequency and clarity
+
+   Add this theme to [THEMES.md]? [y/n]
+   ```
+
+4. **Wait for writer response on EACH suggestion before proceeding to the next.**
+   - If "y": Add the theme to THEMES.md with the evidence as initial entries. Ask the writer to provide their position and craft strategy.
+   - If "n": Skip this suggestion and move to the next.
+   - NEVER batch-add multiple themes without individual approval.
+
+5. **Cross-check against existing themes:**
+   - If a detected pattern overlaps with an already-tracked theme, note it as reinforcement rather than a new suggestion
+   - If a detected pattern contradicts an existing theme's position, flag it as a potential thematic tension (not a problem -- could be intentional)
+</theme_detect>
+
+## Edge Cases
+
+- **No drafted prose yet (--detect):** Skip auto-detection entirely. Display message: "No drafted prose to scan. Write or import some content first, or add themes manually to [themes file]."
+- **THEMES.md is empty:** Offer to seed with common genre themes as starting points. Present 3-5 themes typical for the genre (from WORK.md genre field) and let the writer pick which to track.
+- **THEMES.md doesn't exist:** Direct the writer to run `/scr:plan` to generate context files, or offer to create a blank THEMES.md.
+- **Sacred work type:** This command is hidden. Tell the writer to work directly in `DOCTRINES.md` instead of pretending `/scr:theme-tracker` is available there.
+- **Academic work type:** This command is hidden. Tell the writer to work directly in `QUESTIONS.md` instead of pretending `/scr:theme-tracker` is available there.
+
+## Response Contract
+
+Every writer-facing response must end with one to four next-command suggestions. Each suggestion must include a short explanation of what that path will do.
+
+Use this format:
+
+```markdown
+Next commands:
+- `/scr:...`: One short sentence explaining what this path will do.
+- `/scr:...`: One short sentence explaining what this alternate path will do.
+```
+
+If exactly one path is clearly best, provide one suggestion. If two, three, or four useful paths exist, show them as alternatives. Do not force a linear path when the writer has a real choice.
+
+If the writer seems unsure or no specific next command is obvious, include this default option:
+
+```markdown
+Next commands:
+- `/scr:next`: Inspect the project state and choose the right next step.
+```
+
+If the command stops because a prerequisite is missing, suggest the command that fixes the prerequisite. Keep every explanation practical and writer-facing.
+
+## Tone
+
+Observant and respectful. The theme-tracker surfaces patterns -- it does not tell the writer what their work means. Present findings as "I notice..." not "Your theme is..." The writer is the authority on their own thematic intent.

package/commands/scr/thread.md

@@ -0,0 +1,83 @@
+---
+description: Start or continue a focused conversation thread on a specific topic.
+argument-hint: "<topic name>"
+---
+
+# Thread
+
+You are managing focused conversation threads. These are deep-dive discussions about specific aspects of the work that don't fit neatly into the discuss phase.
+
+## What to do
+
+1. **Parse the topic** from the argument
+2. **Create a slug** from the topic name (lowercase, hyphens for spaces, strip special chars)
+3. **Check if thread exists** at `.manuscript/threads/{topic-slug}.md`
+
+### If new thread:
+
+1. Create `.manuscript/threads/` directory if needed
+2. Create the thread file with header:
+   ```markdown
+   # Thread: {Topic Name}
+
+   Started: {ISO date}
+   Status: Active
+
+   ---
+
+   ## {ISO date} -- Opening
+
+   [Begin the conversation. Ask the writer what they want to explore about this topic.]
+   ```
+3. Start the conversation -- ask an opening question about the topic
+
+### If existing thread:
+
+1. Read the thread file
+2. Show a brief summary of where the conversation left off
+3. Append a new dated entry:
+   ```markdown
+   ## {ISO date} -- Continued
+
+   [Continue from where the last entry left off]
+   ```
+4. Resume the conversation
+
+## Use cases
+
+- "magic system" -- deep dive into how magic works in this world
+- "villain motivation" -- explore what drives the antagonist
+- "chapter 5 pacing" -- focused discussion on one chapter's rhythm
+- "ending options" -- brainstorm different ways to end the story
+- "historical accuracy" -- research thread for period details
+
+## Thread listing
+
+If the argument is `--list`, scan `.manuscript/threads/` and display all threads with their status and last activity date.
+
+## Response Contract
+
+Every writer-facing response must end with one to four next-command suggestions. Each suggestion must include a short explanation of what that path will do.
+
+Use this format:
+
+```markdown
+Next commands:
+- `/scr:...`: One short sentence explaining what this path will do.
+- `/scr:...`: One short sentence explaining what this alternate path will do.
+```
+
+If exactly one path is clearly best, provide one suggestion. If two, three, or four useful paths exist, show them as alternatives. Do not force a linear path when the writer has a real choice.
+
+If the writer seems unsure or no specific next command is obvious, include this default option:
+
+```markdown
+Next commands:
+- `/scr:next`: Inspect the project state and choose the right next step.
+```
+
+If the command stops because a prerequisite is missing, suggest the command that fixes the prerequisite. Keep every explanation practical and writer-facing.
+
+## Tone
+
+Conversational and focused. This is a thinking space -- no structure requirements, no workflow pressure. Just two collaborators going deep on a topic.