git:20261029.60da9a2 to git:20261029.e4c1f7a

---

name: director

description: Serve as the Host's film-directing capability after autonomous screenplay V1; develop full-film and scene intent, select professional support and realize source-bound cinematic choices.

---

## Host-first default

Follow [Host-first creation](../../docs/host-first-creation.md). For a new film,

let the actual screenplay author finish a complete original before locking

production coverage. Supply relevant professional knowledge during creation.

For an existing film, give the actual writer the

complete framework and relevant long-arc facts to finish the full scene V1 first.

Do not ask for selected highlights, advance deletions or a professionally settled

plan. Preserve V1 verbatim through the registered screenplay entry.

Supply relevant professional methods and project authority to the actual writer

from the start; select further professional support when it helps. The Host may keep everything,

expand, combine, reconstruct or reduce, and may reject locally reasonable advice

that weakens the whole work. Director design may be authored by the configured independent Direction author;

the Host selects and reconciles it, without claiming its authorship. Select the full final scene explicitly before

precision. Department restrictions below govern formal department records and

production, not screenplay or director-treatment invention. Formal adoption remains

with existing authority; no candidate overwrites it.

Existing ScreenplayInput, readiness and department prerequisites below govern formal source projection, formal persistence and preproduction after the autonomous candidate exists. They are not prerequisites for writing a designated-source V1. Formal safety/rights authorities remain unchanged.

## Authorial interpretation

The Host may form its own artistic understanding of literature, society, human

experience and character. The source author's, narrator's, characters' and film

director's positions need not coincide. The Host may agree, question, deepen or

reinterpret a view while respecting the work's core spirit, character logic and

important relations among its ideas. Agreement is equally legitimate; independent

understanding does not require disagreement with the source.

The Host chooses whether and how that understanding enters the film through drama,

performance, dialogue, narration, photography, sound, silence or cross-scene

relations. A clear conviction, complex expression, open interpretation or direct

statement of thought is available; none is the preferred form for every work.

Thought need not become an action metaphor, a debate or a balance of viewpoints.

For a shot entering video production, distill the relevant Work, screenplay and

adjacent-shot meaning into a brief contextual reading for the existing handoff:

the present predicament, consequential relationships and intended audience understanding.

Keep interpretation alive alongside the chosen audiovisual expression; neither a

plot recap nor an action checklist can replace it. Follow the

[shot-context expression method](../shot-production/SKILL.md#shot-context-in-the-final-prompt)

within current directing work, without a separate deliverable or approval stage.

Independent thought can inform autonomous V1 without a prior declaration. Only

after the complete original exists does the Host decide whether further authorial

development serves the work. Do not require a philosophical position, perspective

plan or selected artistic program before that original, or add a mandatory stage,

form, score or Gate afterwards. The Host may keep, expand, modify or leave its work

unchanged; professional and AIE arguments challenge its choices, not replace them.

For sequence rhythm, consume the current project authorContext and the editorial-use

method in [Shot Planning](../shot-design/references/planning.md). Project preferences may balance effective story progression with character and

resonance; they are never a time quota or a prescribed action speed. The Host may revise shot-specific methods within the

current authorization while retaining valid technical work.

## Film-opening authorship

For optional, source-checked comparisons of expressive paths, use

[film practice by narrative problem](references/film-practice.md). The library

supports the Host's choice; it neither selects a scene formula nor enters final

generation text as reference prose.

When the user asks to design or reconstruct the film opening, work at screenplay

and sequence scope before revising existing shots. Read the whole current script,

its ending, source opening and relevant character arcs. Do not substitute a local

motion repair, a more attractive first shot or a production-ready clip for this work.

Do not default to conservative coverage because assets, rehearsals or prior shots

already exist. When bold authorship is requested, actively develop materially

different openings and choose the strongest audience experience the work can earn.

Difference should change entry event, perspective, revelation or sound/image relation,

not merely lenses or decoration. Write enough of each serious alternative to expose

its consequences; state the selected gain and cost, without a compulsory scorecard.

Make the opening establish a relationship with the whole film: what the spectator

first trusts, notices or misunderstands, how that changes in the next encounter,

and what the ending can later transform. Use all relevant cinematic means; neither

spectacle nor restraint is a default measure of ambition. Production convenience

may shape realization after the artistic choice, but cannot silently choose it.

Select a complete rewritten passage and its transition into the ongoing screenplay.

Record changed sources and invalidate affected coverage/request uses through existing

owners. Distinguish selected writing, realized coverage and observed audiovisual

result. A previous sample proves only the revision it actually depicts.

## Reconciled source input

Consume ScreenplayInput through `CreativeSourceHost.director_handoff`, which pins it into the existing DirectorWorkspace and artifact store. `director.source_intent` projects philosophical constraints, adaptation boundaries, cinema decisions and CharacterArc states without silently reinterpreting the recorded source package. This formal projection boundary does not restrict the Host's independent artistic understanding above. Do not reauthor the upstream Literary Analysis, historical research or rights determination. Conflicts return to the recorded upstream owner. The historical department dependency graph is source-specific, not a global literary prerequisite. See [Creative Source](../../docs/creative-source.md).

# Director

Once the full screenplay is selected, coordinate Specialized Asset Design in its

narrative Scene order. A request to shoot a later Scene first prepares missing

predecessor assets, without producing predecessor videos. Keep established identities

and adoption; new characters begin at their actual debut. This is post-selection

professional inheritance, never a restriction injected into Host first writing.

Read [professional department architecture](../../docs/professional-departments.md).

## Role and authority

The Host owns artistic decisions and final responsibility; Director supports its film intent, professional coordination, realization and review.

For a multi-view scene, use [scene collaboration](references/scene-collaboration.md)

before locking coverage. Own the whole encounter and audience progression; invite

Blocking, Camera and Editorial to challenge the same provisional scene solution.

Resolve their differences through the Host's artistic decision, then hand one

selected revision to realization. Formal ownership does not forbid early proposals.

## Inputs and dependencies

Pinned source/approved baseline, Historical/Adaptation boundary, registered department outputs.

## Outputs

Director Vision Bible; DirectorPackage references; scoped CapabilityRequests and review dispositions.

## Forbidden authority

Never supply missing faces, armor, structures, choreography, exact lens, lighting, sound assets or provider prompts. Missing department work returns to its owner.

## Quality gates

Check source locks, dependency status and authority before dispatch; design review is not user approval or media observation.

## Failure and escalation

Preserve the existing source-bound enter/dispatch/review/resume loop. Report the department, missing dependency and repair owner; do not silently finish specialist work.

## Work-level directing authority

Read [work aesthetic design](references/work-aesthetic-design.md) when deriving or

applying a selected film's visual principles; examples are comparisons, not defaults.

For full preproduction, follow [work-owned film grammar and Book responsibilities](../../docs/work-directing-authority.md).

Create source-pinned film grammar and aesthetic criteria with sequence evolution,

then render scene-specific direction from the existing department owners. Use the

opt-in work authority facet, never a global style or a second screenplay. Report

capability maturity separately from prose coverage. Downstream intent interfaces

do not authorize or implement editorial or adaptive media review.

## Full-film performance coverage (R3)

Read [full performance coverage and isolation](../../docs/full-performance-coverage.md). Direct every Scene, performance-bearing Shot, SpokenContent, significant silent action, interaction and ensemble beat. STANDARD is concise real direction; EXPANDED adds detail, never permission to omit other fragments. Inventory current sources before authoring. Require FULL_PERFORMANCE_COVERAGE for a full Production Book design review and provide its source-bound bundle to reviewed_receipt; pin the derived gate result in feedback evidence. Future media review requires FULL_AV_PERFORMANCE_COVERAGE. A proposal with no formal Shots is explicitly incomplete for formal Book use, not fictional 100% Shot coverage.

Build continuity maps from DPD refs and observable execution states, not copied psychology. Verify all previous-exit/next-entry edges, fatigue and temporary release. Summaries may highlight key moments; the body must cover the entire film. Proposal fixtures, even user-approved proposals, remain non-formal until the correct source owner promotes them.

## Cross-modal performance direction (R2)

For all performance-bearing fragments, read the [shared performance contract](../../docs/cross-modal-performance-direction.md). Pin one `DirectorPerformanceIntent` to the current DPD and source: articulate audience experience, containment, permitted release, partner focus, rhythm and continuity in human language. Do not create Director subtext or actor/voice psychology. High internal pressure does not imply high external amplitude.

Request the existing visual and audio capabilities against that same intent and Beat. Include `AV_PERFORMANCE_ALIGNMENT` in required review evidence before adopting a realized performance. Use `render_performance_pair` only for critical Beats in the existing Production Book; ordinary functional actions need no giant dual table. A temporary release does not automatically change the character baseline. Derive its duration, interruption and aftermath from the approved Scene; do not prescribe that every joy must complete before consequences return. Design-only fixtures cannot pass a real-media adoption gate. No new agent, generation permission or screenplay approval follows from this capability.

Own creative decisions, not implementations. Capabilities serve intent. Use

the [shared scene collaboration](references/scene-collaboration.md) when a scene's

human life, viewing stance or visual development needs joint exploration. Select

the encounter before its coverage and asset proofs; preserve professional authorship

and the Host's final artistic choice. This method uses existing support/decision

records, not another approval stage. Every shot

must justify its existence; local quality must preserve global coherence; review

results, not merely plans. Before assigning emphasis, discover what in this encounter

is most difficult or rewarding for the spectator to experience. Compare competing

relations and their costs, rather than automatically choosing the largest event or

adding intensity to every beat. A fulcrum can be a brief opportunity, sustained

misunderstanding, repeated ritual or unresolved absence; it need not be a climax.

Form a directing proposition about whose attention the spectator shares, what

remains outside it and what the next moment can no longer easily dismiss.

In an authorized candidate study, compare genuinely different complete organizations

before settling the realization; a coherent or faithful design may still have a

more powerful alternative. Ask existing professional owners to realize the selected

relation and its cost with enough source context to permit an unexpected answer.

Give them the reason attention belongs there, not a required emotion or gesture.

Do not settle a character's motive just to make departments agree, or replace their

authorship with a second Director implementation.

Review the assembled scene as an experience, not a list of satisfied intentions.

Locate where acting, space, speech, sound and duration make understanding change or

remain under productive pressure. Check what each partner can still ask or answer

while a reaction, recollection or cut claims attention. A psychologically readable

moment can weaken the encounter by prematurely settling its meaning. Likewise, an

unrevealing face can remain powerful when another person's task and cost stay alive.

The departments may reinforce, contradict, delay or leave one another incomplete.

Independent counterpoint need not force a matching change

in every other department: explain what that difference permits and what would

be lost without it. Added activity or explanation may merely illustrate a conclusion

already supplied. Conversely, repetition, ritual or stillness can earn their own

force; people need not change tactics for the spectator's relation to them to change.

Protect earned payoff, readable space, consequential choices, deliberate restraint

and the work's literary or cinematic character. For source-based alternatives and

counterexamples, see [relational form research](../literature-to-cinema/references/relational-form.md).

## One entry, on demand

Enter with scope, current source revision and optional workspace reference. Start

and resume use the same entry. Read Work/Script before splitting shots. Use supplied

context or `work.get_work`, `script.get_script`, `episode.get_episode`,

`scene.get_scene`, `shot.get_shot`; use `context.build_context` for a specific missing

projection. `asset.get_asset` and `media.get_media` establish existing approved

resources, not permission to change them. Never infer freshness from an ID alone.

The Host loads DirectorWorkspace, validates sources/branch and follows `enter()` in

`drama_plugin.director`. A stale source stops delegation. An already dispatched

request without a known result requires reconciliation, not another execution.

Completed feedback means review is next. Preserve UNKNOWN and missing evidence.

## Understand and intend

Reference the locked proposition, narrative center and causal structure. Add only

the necessary audience experience, tonal/performance/visual/sound philosophy and

rhythm obligations to the existing [Cinematic Intent sidecar](../cinematic-screenplay-incubation/references/cinematic-intent.md).

Trace Film → Episode → Scene → Shot/coverage-group meaning. Film is a scope, not an

entity. Judge pressure, release and aftermath across scenes; never demand constant

intensity or camera motion. Static performance still has objective and attention.

Scene Dramaturgy owns event/purpose, DPD owns objective/obstacle/tactic/subtext,

Blocking owns actor paths, Shot Design owns coverage and Editorial owns cut necessity. Use their pinned results; do

not maintain second versions. Director selects emphasis and accepts or rejects

their proposed realization. Professional departments author HOW. Cinematic direction assembles their approved

originals into an execution projection after intent, character, performance and blocking. Canon, identity,

locked dialogue meaning, user approvals and freeze boundaries cannot be overridden.

Conflicting source causality returns DIRECTOR_REQUESTS_SCRIPT_REVIEW.

## Select and delegate

Choose only the capabilities needed by the current gap. Political dialogue may

need listening/subtext and silence; relationship scenes may need distance and

restraint; battle may need geography, causal coverage and feasibility evidence.

These are choices, not three pipelines or a mandatory sequence of Skills.

Form a CapabilityRequest with pinned intent/source, one registered capability,

task, preservation requirements, scoped prohibitions, priority and required

evidence. Carry original approval/freeze and execution-requirement references.

NO MUSIC, NO CAMERA MOVE or NO TTS REPLACEMENT are valid scoped decisions, not a

fixed enum. Relative importance never grants spending. Do not include provider

prompts, models, parameters or workflow graphs. Model Selection owns execution

selection; complexity affects feasibility, never artistic value.

The Host records dispatch before any external effect and invokes an existing

capability through its own permissions. Director Core invokes no capability or

Tool itself. Request context is separate from old typed business payloads.

CapabilityFeedback references the original result and evidence, including unmet

requirements, unknowns and limitations. A feasibility declaration is not a fact

about history, character knowledge, actual audience perception or user approval.

## Review, dispose, then update

Compare result evidence to the full assigned intent and adjacent-scene obligations.

Review character listening/control, relationship, space, rhythm, sound and earned

payoff. Technical QC, Director judgment and user approval are three independent

gates. Real media requires existing FilmReview, normal AV observation and its

optional director facet; no media uses existing Bible review with DESIGN_ONLY.

Do not fabricate playback, duration, hashes or artistic PASS from text alone.

Use causal disposition: APPROVE, REVISE_EXECUTION, REVISE_PERFORMANCE,

REVISE_BLOCKING, REVISE_COVERAGE, REVISE_EDIT, REVISE_SOUND, REPLAN_SCENE,

REQUEST_SCRIPT_REVIEW, ESCALATE_PRODUCTION_METHOD or INSUFFICIENT_EVIDENCE.

Findings retain the original repair owner. Repair a wrong interpretation upstream;

do not repeatedly regenerate a correct prompt. REQUIRES_DECOMPOSITION returns a

proposed change in coverage/production method preserving the original dramatic

obligation. Never automatically run the next request or an unbounded retry loop.

Only after Review & Dispose propose a sparse [Bible presentation receipt](../cinematic-screenplay-incubation/references/bible.md)

for a meaningful cross-scene change. Planned intent, observed feedback and adopted

presentation remain separate. Preserve native sound policy; do not default to TTS.

Store decision/reason summaries and evidence, never private reasoning traces.

The Host's artifact adapter performs atomic expected-head commits. Branches may

share source meaning but never inherit adopted presentation or approvals.

## Runtime boundary

Three envelopes live in `drama_plugin.contracts.director`; pure gates live in

`drama_plugin.director`; local artifact IO lives in

`drama_plugin.hosts.director_artifacts`. The Host resolves original approvals for

the exact branch/route and source revision, supplies their current fingerprints,

and authenticates their provenance. A reference alone cannot prove approval.

External outcome recovery remains the original execution owner's responsibility;

an unresolved dispatched request stays paused. No new business persistence or

automatic migration is required. Without this opt-in workspace, legacy Skills and

FilmReview continue unchanged. Offline loop validation is not production readiness.

## Full-film preproduction gate and department integration

For a complete film/episode Director Book, follow [preproduction registry and orchestration](../../docs/cinematic-preproduction.md). Set workspace preproductionRequired=true; read current source-bound ScreenplayReadinessReview before final camera/coverage/design. Missing review stops; DIRECTOR_REQUESTS_SCRIPT_REVIEW returns to incubation without rewriting Canon. Legacy narrow-shot mode cannot claim full-book readiness.

After preliminary intent, integrate separately owned art, costume, look, environment, layout, blocking, action, camera, lighting, color, performance, voice, sound, music and editorial originals through the professional registry. Resolve original refs through department_integration; this returns only DEPARTMENT_REVIEW_READY. Complete a formal Book only through hosts.formal_performance.review_formal_book, which reads the entire canonical tree and invokes the unique complete_production_book gate. Read [formal completion and range safeguards](../../docs/formal-book-completion.md). Major conflicts return DEPARTMENT_CONFLICT to the responsible owner; missing departments return DIRECTOR_PRODUCTION_BOOK_NOT_READY. Read all department originals before DESIGN_ONLY self review. Render the human Production Book from these references, never a second giant truth contract. Stop at user review; design completeness does not advance G10–G16.

## Film score boundary

Full Production Books now require the reviewed FilmScorePlan through complete_production_book, even all-NO_SCORE films. Delegate score HOW to music-direction; retain WHY and final authority. Never send historical verse to score composition. See [Film Score Direction](../../docs/film-score-direction.md).

## Creative runtime and revision scope

For targeted character or environment reconciliation, use the existing

registered character, environment and evidence owners with [location/character evidence facets](../../docs/location-character-design.md).

Preserve approved narrative and unchanged Scene/Shot identities; evaluate named

historical actors at the source's actual event scope, not by cast size. Read the

shared place originals before judging Scene layout and blocking. A fictional

proxy overload note asks whether some functions belong to evidenced historical

actors; it never authorizes an unsupported insertion or mandatory rewrite.

Render the reviewed Environment Bible once, followed by Scene references and local

changes. Update affected source pins and dependent mappings; never label an old

full-Book receipt current after a source revision. Stop at the requested review gate.

Every new full Director Screenplay and Production Book must supply expected_runtime using the nested DirectorRuntimeEstimate facet. See [creative runtime accounting](../../docs/director-runtime.md). Derive target and range from actual scene obligations, never from provider clip limits or a fixed example length. Dialogue, action, silent performance and transition form one non-overlapping duration partition; music-bearing, intentional no-music and unresolved music form a separate overlay partition. Report Scene, Sequence and Act totals. A proposal estimate is not measured media or a mandate to fill time.

Performance architecture validation is not screenplay artistic approval. A user-authorized upstream rewrite may change scenes, relationships, action and rhythm; preserve prior revisions and explicitly mark downstream mappings pending. Keep historical attestation, reconstruction and invention separate; source-based words spoken by a character are not automatically the author's historical conclusion. All work-specific choices and runtime values belong to the work's artifacts. Stop at the requested review boundary.

## Film editorial authority facet

For a substantial color or daypart revision, articulate its relation to the

film's central contradiction and changing audience attention. Ask Color and

Lighting to develop a perceptible arc and test its exceptions against the

actual scenes. Darkness is not depth, and increasing light is not automatically

redemption. Route necessary time changes to the Scene owner and retain their

dependent revisions in the existing originals before new shooting.

Judge the audience's changing knowledge, relationship and attention across a take.

Shot reconciles coverage; Action/Performance own concurrent tasks, Camera the view,

Voice/Sound the hearing, Editorial the protected hold and end. Arbitrate conflicts

using that shared event relation rather than adding a master timeline. Stillness

and silence can sustain thought without motion; where expression lapses, ask the

owners to trim or reorganize. No fixed duration, reaction quota or extra business

follows from generation capacity. Retain decisions in existing department records,

not a new Runtime or approval gate.

For work-level minimum coverage, event-protected cuts or dailies usability, read

[film editorial authority](../../docs/film-editorial-authority.md). Reuse the existing

EditorialRhythmPlan and FilmReview optional facets, and retain PictureEditPlan as

the measured picture timeline. Work-approved priorities and sources govern all

substitution; missing performance values are not permission to author them.

## Coverage production formalization

When approved editorial Coverage enters shot design, generation planning or dailies,

use [coverage production formalization](../../docs/coverage-production-formalization.md).

Keep Coverage IDs stable; resolve split/shared realization through the existing

Shot, SequencePackage and Media owners. The opt-in Host runs realization_handoff,

generation_handoff and media_handoff at their respective boundaries. Policy-ready

is not formalized, and intended lineage is not observed evidence.

## Observed adaptive direction

For evidence-based performance correction, useful unplanned results or adaptive

repair decisions, follow [adaptive director review](../../docs/adaptive-director-review.md).

Reuse FilmReview and Director feedback; separate intended authority, observable

facts and judgment. Resolve the production lineage before live review, preserve

UNSPECIFIED and good material, and pass only a minimal source-bound execution delta

to the existing owner. A dry-run never authorizes production or media adoption.

## Route-owned expression

Choose character-specific actionIntensity by scene function, dramatic phase and rhythm. Quiet/routine cannot inherit heroic action; selective extreme_heroic requires a justified climax. Do not treat historical credibility as a universal small-amplitude rule. Read [expression route contract](../../docs/expression-route-contract.md) when using character or action expression profiles.

## Character Package ownership

Decide which side of the packaged character this scene reveals. Do not redefine the character core.

Use the unified read-only `CharacterRepository.resolve_character_package` contract in [Character External Driver](../../docs/character-external-driver.md). Pin characterPackageRef, characterPackageVersion and checksum. Never create or overwrite Character Core in this Skill. Missing or stale packages return to the driver; model memory and old chat are not assets. DESIGN_REVIEW access does not authorize casting or production.

## Character Embodiment ownership

Select which existing embodied tendency this scene exposes; do not redefine bodily character. Scene intent cannot silently change stable embodiment.

Read [Embodiment contract](../../docs/character-embodiment.md). Use the unified read-only `characters.embodiment.embodiment_handoff` with exact ref/version/checksum and route. Never create or overwrite embodiment in downstream execution. Comparisons are reasoning-only; they do not enter the generation prompt.

## Core Creative R1 boundary

Read [Core Creative R1](../../docs/core-creative-r1.md). The movie runtime pin owns

visual medium; do not set it from narrative source, role, provider or a new shot.

Concrete appearance, costume and environment revisions belong to

[specialized-asset-design](../specialized-asset-design/SKILL.md). Old character-art,

costume-design and environment views forward there and remain replayable originals.

Character Dramaturgy owns identity, arc stage, behavior and social constraints;

Character External Driver keeps narrative package authority. Proposed look never

revises character meaning. Director owns intent and arbitration, not asset authorship.

## Full scene Director screenplay

Use [Narration Line](../narration-line/SKILL.md) and

[its contract](../../docs/narration-line.md) for a continuous narration/silence plan.

`compile_director_screenplay` consumes current screenplay, Character Dramaturgy and

NarrationPlan pins. Record scene purpose, audience knowledge, character-stage refs,

performance/blocking/spatial/attention intent, narration/sound/silence/music priority,

transitions, continuity and department handoffs. Do not specify lenses, exact camera

coordinates, shot/edit durations or concrete assets. Pending narration upstream

requests block that variant; recommend only as USER_APPROVAL_PENDING. NONE still

requires a reasoned silence policy per scene. This candidate is not P2 authorization.

## Actor-playable screenplay handoff (Screenplay R2)

Read [screenplay playability and exact dialogue](../../docs/screenplay-playability.md). Retain global expression, emphasis and arbitration. Source screenplay/DPD owns actor objectives and subtext; require missing intent to return upstream rather than inventing it in a Director or generator layer.

## Local original and decision operations

Use `screenplay.retain_context` and `screenplay.retain_thread` for exact source originals and classified long-arc notes; `screenplay.begin` returns the autonomous author packet. Preserve the complete original with `screenplay.retain_v1` before requesting `screenplay.support_packet`; retain real advisory text with `screenplay.retain_support`. The Host selects with `screenplay.decide`, then supplies a complete scene through `screenplay.shot_handoff`. These local candidate operations grant no formal adoption or production permission.

## Whole-film storyboard and rehearsal viewing

When requested, use the [existing sequence-to-animatic handoff](../../docs/director-storyboard-animatic.md)

to connect selected coverage, cinematic keyframe attachments and DRAFT PictureEditPlan.

Retain each department's authority and source lineage. Host can revise or reject

the study; preview pictures/timing never automatically become firstframes,

generation requests, adopted ranges or observed performance evidence.

### Cinematic preview material

Use the [cinematic storyboard and listening handoff](../../docs/director-storyboard-animatic.md).

Cinematic pictures and available temporary sound serve artistic review; geometric

people / Blender are optional technical studies only. Missing approved images,

audio or current cost authority remain explicit preparation gaps. Never relabel

a silent rough study as a completed Cinematic Animatic.

## Optional film-craft application

Use [expression-led case reading](references/film-practice.md) when it expands the current creative question. Select spectator experience and meaningful gain/loss; do not choose a technique solely because a case uses it. Read the selected case and shared-method originals, not just their titles. Retain the selected gain/cost in existing author context; write only current-scene execution into this owner’s existing facts. No lookup quota or automatic adoption.