kbpd.ai — home Talk to us
← Proof Studio

Persona council · Demonstrator

A council run beats one generic lens

A team is facing a hard product-planning decision and needs architecture, journey, verification, operations, and business judgment separated before synthesis.

demo launch candidate public summary sanitized 10 evidence rows 14 evidence values
Inspect evidence Evidence JSON

council fanout

Persona council fanout

Sanitized projection
The graph shows the marketable behavior: one brief enters, five specialist lenses return different evidence, and the synthesis keeps the lanes visible.

HERMES case-study frame

Read the case before the evidence room.

A reviewer should be able to answer seven things before opening the JSON export: what decision is in front of them, what changed against a cold start, which row carries the claim, where the value appears, what source trail exists, what purchase decision it informs, and what the case does not prove.

  1. 01 · Buyer question

    What is being decided?

    A team is facing a hard product-planning decision and needs architecture, journey, verification, operations, and business judgment separated before synthesis.

  2. 02 · Cold start

    Single assistant response

    A generic assistant can produce a plausible plan from the prompt alone, but the reasoning lanes, source assumptions, and evidence standards are fused together.

  3. 03 · KBPD-backed

    Council fanout

    The same hard brief is routed through specialist lenses, prompt archive patterns, and retained project source before one synthesis is written.

  4. 04 · Boundary

    Demonstrator

    This is one sanitized run, not a statistical benchmark.

Best public row five-lanes-separated Publicly inspectable

Proof row

Five lanes are separated before synthesis

The public projection separates architecture, journey, verification, operations, and business lanes before the synthesized recommendation.

Supports the claim that the council run is more inspectable than one blended assistant response.
Evidence value
Specialist lanes: 5 Architecture, journey, verification, operations, and business lenses are shown separately.
Source trail
Council projection sanitized diagram · Public projection
Buyer decision
Compare the role samples The visible difference is auditability. The council run lets a reviewer inspect which role, source class, and corpus pattern found which constraint before accepting the synthesis.
Limit
Demonstrator This is one sanitized run, not a statistical benchmark.

Evidence room

Inspect the evidence, values, and limits.

This is the public review surface for the case. Each row names the observation, the evidence values it stands on, the source trail, the boundary, and the matching structured entry in the evidence export.

Evidence JSON
Review order Question -> evidence -> value -> source -> limit Start with the case question, then inspect what the claim is allowed to stand on.
Evidence export /proof/persona-council-fanout/evidence.json The same evidence rows are emitted as JSON for independent review.
Claim boundary Demonstrator The page cannot claim more than this status supports.
Source availability

A source trail is not always a raw public artifact. Public projection means the source has been redrawn or summarized; retained source means it is named for audit but withheld until redaction clears.

Public projection
6
Retained source
4
Pending run
0

Public evidence excerpts

Readable snippets tied to evidence rows

These excerpts are the public bridge between retained raw sources and the evidence table below. Each one says what was preserved, what was redacted, and which evidence rows it supports.

  1. Sanitized reconstruction five-lane-council-excerpt

    Five-lane council excerpt

    One complex brief is routed to architecture, journey, verification, operations, and business lanes before synthesis. Each lane has a distinct evidence job.
    Source label
    Sanitized transcript storyboard
    Redaction note
    Redrawn from private captures; source screenshots, issue keys, workspace paths, and client identifiers are not exposed.
    Supports evidence rows

    five-lanes-separated

  2. Sanitized reconstruction corpus-source-roster-excerpt

    Corpus source roster excerpt

    The public mechanism names four source classes: role roster, prompt archive patterns, prior decisions, and project source.
    Source label
    Corpus-backed council mechanism
    Redaction note
    The excerpt preserves the source-class structure while withholding the underlying prompt archive and raw lane outputs.
    Supports evidence rows

    raw-screenshots-not-publicfour-corpus-source-classes

  3. Sanitized reconstruction t3-proof-row-excerpt

    T3 proof-row excerpt

    Five rows compare the cold baseline against specialist lanes: architecture, journey, verification, operations, and business each carry a different evidence signal before synthesis.
    Source label
    Public T3 proof-row table
    Redaction note
    The table shows source classes and public signals only; source prompt text, private captures, and unredacted lane outputs stay retained.
    Supports evidence rows

    cold-baseline-blends-lanest3-proof-row-table-publishedhermes-janus-language-gate-visible

  4. Sanitized reconstruction cold-baseline-review-board-excerpt

    Cold baseline review board excerpt

    The cold baseline is shown as one blended recommendation beside the KBPD-backed artifact, then tested across five checks: role boundary, source standard, claim pressure, public proof path, and language gate.
    Source label
    Public cold baseline review board
    Redaction note
    The board preserves the review structure without exposing raw prompt text, private screenshots, workspace paths, issue keys, account details, or unredacted lane outputs.
    Supports evidence rows

    cold-baseline-blends-lanescold-baseline-review-board-publishedt3-proof-row-table-published

  5. Sanitized reconstruction color-separation-lane-board-excerpt

    Color-separation lane board excerpt

    Five persona lanes are redrawn as overlapping color-separation plates. Each lane keeps a source class and evidence job visible before the public artifact resolves.
    Source label
    Public color-separated proof board
    Redaction note
    The board is a public projection of the retained council mechanism; raw screenshots, private source text, workspace paths, account details, issue keys, and unredacted lane outputs remain withheld.
    Supports evidence rows

    color-separation-lane-board-publishedfive-lanes-separatedraw-screenshots-not-public

  6. Sanitized reconstruction sanitized-run-gallery-excerpt

    Sanitized run gallery excerpt

    Five public frames replace the private screenshot stack: brief arrives, corpus attaches, lanes run separately, synthesis keeps the trail, and the buyer gets a review path.
    Source label
    Public run gallery
    Redaction note
    The gallery preserves run structure while withholding raw prompt text, workspace paths, issue keys, private captures, and unredacted lane outputs.
    Supports evidence rows

    raw-screenshots-not-publicsanitized-run-gallery-published

  7. Sanitized reconstruction capture-review-board-excerpt

    Capture review board excerpt

    Five review rows map private capture classes into public frames: panel launch, architecture lens, journey and verification lenses, operations lens, and business-language gate.
    Source label
    Public capture review board
    Redaction note
    The board names retained signal classes and redaction boundaries while withholding raw screenshots, workspace paths, account details, operational URLs, issue keys, and unredacted lane outputs.
    Supports evidence rows

    capture-review-board-publishedraw-screenshots-not-public

Evidence rows
10
Reviewed public observations
Evidence values
14
Metric and run values exposed for review
Source trail
10
Artifacts named for review
Claim status
Demonstrator
Claim boundary for publication

Claim under review

Role-specific context and a retained prompt corpus produce more inspectable work than a single generic assistant because each lane has a different evidence job before synthesis.

The visible difference is auditability. The council run lets a reviewer inspect which role, source class, and corpus pattern found which constraint before accepting the synthesis.

Date
June 2026
Sample
One sanitized council run
Provenance
Public projection of a retained multi-lane planning run
Reproducibility
The public projection is reproducible as a proof pattern; the retained source requires redaction before publication.
01 Five lanes are separated before synthesis Publicly inspectable evidence_facts[0]
Observed
The public projection separates architecture, journey, verification, operations, and business lanes before the synthesized recommendation.
Evidence
Council fanout visual, transcript storyboard, and mechanism stage each show five named lanes converging only after lane outputs exist.
Supports
Supports the claim that the council run is more inspectable than one blended assistant response.
Evidence values used
Source trail
Limit that constrains this row

This is one sanitized run, not a statistical benchmark.

Evidence id five-lanes-separated Review URL #fact-five-lanes-separated
02 Color-separation lane board is public Publicly inspectable evidence_facts[1]
Observed
The public persona page now shows five persona lenses as layered color-separation plates that converge into one resolved buyer artifact.
Evidence
The `Color-separated proof` board links the same retained corpus source into Hephaestus, Artemis, Cassandra, Mnemosyne, and Argus/HERMES lanes, then resolves them into a decision, limitation, and buyer-action artifact.
Supports
Supports the requirement that the persona proof be graphical and brand-native, not only a text table or JSON packet.
Evidence values used
Source trail
Limit that constrains this row

The source prompt and lane outputs are not public until redaction is complete.

Evidence id color-separation-lane-board-published Review URL #fact-color-separation-lane-board-published
03 Cold baseline blends the evidence jobs Publicly inspectable evidence_facts[2]
Observed
The public case describes the cold start as one recommendation where architecture, journey, verification, operations, and business judgments are fused together.
Evidence
Before artifact, storyboard step 01, and the T3 proof-row table all name the cold baseline as one blended response without separate source standards.
Supports
Supports the contrast that the council run is easier to review because it separates the evidence jobs before synthesis.
Evidence values used
Source trail
Limit that constrains this row

The run gallery is a redrawn projection, not raw screenshot evidence.

Evidence id cold-baseline-blends-lanes Review URL #fact-cold-baseline-blends-lanes
04 T3 proof rows are visible before the transcript Publicly inspectable evidence_facts[3]
Observed
The public persona page shows a five-row T3 table before asking the reader to accept the broader claim.
Evidence
The T3 table lists each lane, the cold-start failure mode, the corpus-backed signal, source class, and buyer decision in one visual review surface.
Supports
Supports the requirement that the proof case lead with reviewable rows instead of only narrative or JSON.
Evidence values used
Source trail
Limit that constrains this row

The capture review board is a publication map, not the underlying private capture set.

Evidence id t3-proof-row-table-published Review URL #fact-t3-proof-row-table-published
05 Cold baseline review board is public Publicly inspectable evidence_facts[4]
Observed
The public persona page now shows the cold artifact beside the KBPD-backed artifact and lists five baseline checks before the T3 table.
Evidence
The `Cold baseline artifact` section includes role boundary, source standard, claim pressure, public proof path, and language gate checks, each with an evidence-row link.
Supports
Supports the requirement that the cold-start contrast be visible as an artifact review surface, not only described in prose.
Evidence values used
Source trail
Limit that constrains this row

The public page demonstrates reviewability and role separation, not automatic correctness.

Evidence id cold-baseline-review-board-published Review URL #fact-cold-baseline-review-board-published
06 Raw screenshots are intentionally withheld Publicly inspectable evidence_facts[5]
Observed
The proof case publishes redrawn mechanisms and explicitly reports zero public source screenshots.
Evidence
Metric row `Public source screenshots: 0`; source artifact list marks retained run captures as private evidence; the run gallery labels itself as a public projection.
Supports
Supports the public/private boundary while explaining why the public page shows projections instead of raw captures.
Evidence values used
Source trail
Limit that constrains this row

This is one sanitized run, not a statistical benchmark.

Evidence id raw-screenshots-not-public Review URL #fact-raw-screenshots-not-public
07 Five-frame sanitized run gallery is public Publicly inspectable evidence_facts[6]
Observed
The public persona page now redraws the retained run as five frames: brief, corpus, lanes, synthesis, and buyer proof path.
Evidence
The `Sanitized run gallery` section shows retained signal, public projection, and buyer read for each frame, plus an evidence-row link.
Supports
Supports the requirement that raw persona run captures get a publication-safe gallery projection before the case becomes a stronger flagship proof.
Evidence values used
Source trail
Limit that constrains this row

This is one sanitized run, not a statistical benchmark.

Evidence id sanitized-run-gallery-published Review URL #fact-sanitized-run-gallery-published
08 Capture review board maps private screens to public proof Publicly inspectable evidence_facts[7]
Observed
The public persona page shows five capture-review rows that connect retained screenshot classes to public frames, redaction boundaries, and exact evidence-row links.
Evidence
The `Capture review board` section includes panel launch, architecture lens, journey/verification lenses, operations lens, and business-language gate rows, each with retained signal, public frame, redaction boundary, and an evidence link.
Supports
Supports the requirement that private persona screenshots become a usable public proof path rather than an unreviewable screenshot dump.
Evidence values used
Source trail
Limit that constrains this row

This is one sanitized run, not a statistical benchmark.

Evidence id capture-review-board-published Review URL #fact-capture-review-board-published
09 Four corpus source classes are represented Publicly inspectable evidence_facts[8]
Observed
The public mechanism shows role roster, prompt archive patterns, prior decisions, and project source as inputs to the run.
Evidence
Metric row lists four corpus sources; source artifact list names role registry and prompt archive sources.
Supports
Supports the retained-corpus claim without exposing private source material.
Evidence values used
Source trail
Limit that constrains this row

This is one sanitized run, not a statistical benchmark.

Evidence id four-corpus-source-classes Review URL #fact-four-corpus-source-classes
10 HERMES and cold-reader gates are named Publicly inspectable evidence_facts[9]
Observed
The public case names the public-language lane and the cold-reader gate as corpus owners before the result becomes buyer-facing copy.
Evidence
Corpus ownership cards and the T3 proof-row table identify HERMES, archive discipline, session memory, and cold-reader review as separate responsibilities.
Supports
Supports the claim that the prompt corpus is managed as a reusable product asset rather than hidden chat history.
Evidence values used
Source trail
Limit that constrains this row

This is one sanitized run, not a statistical benchmark.

Evidence id hermes-janus-language-gate-visible Review URL #fact-hermes-janus-language-gate-visible

Evidence values

The values this proof is allowed to stand on

Field Value Meaning
Specialist lanes Value handle: data-01-specialist-lanes 5 Architecture, journey, verification, operations, and business lenses are shown separately.
Public source screenshots Value handle: data-02-public-source-screenshots 0 The public case is redrawn from sanitized structure; private screenshots remain internal.
Corpus sources Value handle: data-03-corpus-sources 4 Role roster, prompt archive, prior decisions, and project source are represented in the public mechanism.
Claim boundary Value handle: data-04-claim-boundary demo This proves the pattern is inspectable, not that it wins every planning task.
T3 proof rows Value handle: data-05-t3-proof-rows 5 Five public rows connect cold baseline, specialist lane, evidence signal, source class, and buyer decision.
Baseline checks Value handle: data-06-baseline-checks 5 Five public checks make the cold-start artifact reviewable before the reader reaches the T3 table.
Public run projections Value handle: data-07-public-run-projections 5 Five redrawn gallery frames show the retained run structure without publishing raw screenshots.
Capture review rows Value handle: data-08-capture-review-rows 5 Five publication-safe rows map private capture classes to public frames, redaction boundaries, and evidence links.
Color-separated lenses Value handle: data-09-color-separated-lenses 5 Five persona lanes now render as layered color-separation plates before the longer mechanism and transcript surfaces.
Run date Value handle: data-10-run-date June 2026 Date label attached to this public proof case.
Sample Value handle: data-11-sample One sanitized council run Sample size or unit under review.
Provenance Value handle: data-12-provenance Public projection of a retained multi-lane planning run Where the public projection says the evidence came from.
Reproducibility Value handle: data-13-reproducibility The public projection is reproducible as a proof pattern; the retained source requires redaction before publication. What a reviewer can repeat or inspect publicly.
Claim status Value handle: data-14-claim-status demonstrator The maximum strength of the public claim.

Source trail

Where the public proof can point

  • source-01-council-projection Council projection sanitized diagram · public projection
  • source-02-raw-run-captures Raw run captures internal evidence · internal source retained
  • source-03-persona-roster Persona roster role registry · internal source retained
  • source-04-prompt-archive Prompt archive corpus source · internal source retained
  • source-05-prompt-archive-patterns Prompt archive patterns corpus source · internal source retained
  • source-06-t3-proof-row-table T3 proof-row table sanitized review table · public projection
  • source-07-cold-baseline-review-board Cold baseline review board sanitized review table · public projection
  • source-09-capture-review-board Capture review board redacted capture map · public projection
  • source-10-color-separation-lane-board Color-separation lane board sanitized visual proof surface · public projection

Evidence status mix

What kind of support is each row?

Publicly inspectable
10

Metrics

Signals carried by the case

  • 5 Specialist lanes Architecture, journey, verification, operations, and business lenses are shown separately.
  • 0 Public source screenshots The public case is redrawn from sanitized structure; private screenshots remain internal.
  • 4 Corpus sources Role roster, prompt archive, prior decisions, and project source are represented in the public mechanism.
  • demo Claim boundary This proves the pattern is inspectable, not that it wins every planning task.
  • 5 T3 proof rows Five public rows connect cold baseline, specialist lane, evidence signal, source class, and buyer decision.
  • 5 Baseline checks Five public checks make the cold-start artifact reviewable before the reader reaches the T3 table.
  • 5 Public run projections Five redrawn gallery frames show the retained run structure without publishing raw screenshots.
  • 5 Capture review rows Five publication-safe rows map private capture classes to public frames, redaction boundaries, and evidence links.
  • 5 Color-separated lenses Five persona lanes now render as layered color-separation plates before the longer mechanism and transcript surfaces.

Source artifacts

Where the evidence comes from

  • Council projection sanitized diagram · public projection
  • Raw run captures internal evidence · internal source retained
  • Persona roster role registry · internal source retained
  • Prompt archive corpus source · internal source retained
  • Prompt archive patterns corpus source · internal source retained
  • T3 proof-row table sanitized review table · public projection
  • Cold baseline review board sanitized review table · public projection
  • Sanitized run gallery redrawn run projection · public projection
  • Capture review board redacted capture map · public projection
  • Color-separation lane board sanitized visual proof surface · public projection

Case quality review

This case has a cold-review quality score.

This is a cold-review rubric, not a sales claim. It checks whether the case can be understood, inspected, bounded, and acted on.

Overall 5.0 Launch-readable

Cold-audience clarity

5.0

A reviewer gets the situation, contrast, and buyer decision without reading the JSON.

Evidence visibility

5.0

The case exposes proof rows, values, source handles, and excerpts before the JSON export.

Limitation clarity

5.0

The page names what the case proves and what it does not prove.

Visual hierarchy

5.0

The route uses visuals, storyboards, or surface frames to break the proof into inspectable layers.

Purchase motivation

5.0

The proof sends the buyer to an evidence room, artifact link, pricing path, or local run.

Next depth gate

Cold-audience clarity: Tighten the opening latch or before/after language when the case still needs explanation.

Open case evidence

Evidence docket

Review the evidence rows behind the case.

Evidence JSON
Evidence rows
10
Publicly inspectable
10
  1. 01 Publicly inspectable

    Five lanes are separated before synthesis

    Observed
    The public projection separates architecture, journey, verification, operations, and business lanes before the synthesized recommendation.
    Evidence
    Council fanout visual, transcript storyboard, and mechanism stage each show five named lanes converging only after lane outputs exist.
    Supports
    Supports the claim that the council run is more inspectable than one blended assistant response.
  2. 02 Publicly inspectable

    Color-separation lane board is public

    Observed
    The public persona page now shows five persona lenses as layered color-separation plates that converge into one resolved buyer artifact.
    Evidence
    The `Color-separated proof` board links the same retained corpus source into Hephaestus, Artemis, Cassandra, Mnemosyne, and Argus/HERMES lanes, then resolves them into a decision, limitation, and buyer-action artifact.
    Supports
    Supports the requirement that the persona proof be graphical and brand-native, not only a text table or JSON packet.
  3. 03 Publicly inspectable

    Cold baseline blends the evidence jobs

    Observed
    The public case describes the cold start as one recommendation where architecture, journey, verification, operations, and business judgments are fused together.
    Evidence
    Before artifact, storyboard step 01, and the T3 proof-row table all name the cold baseline as one blended response without separate source standards.
    Supports
    Supports the contrast that the council run is easier to review because it separates the evidence jobs before synthesis.
  4. 04 Publicly inspectable

    T3 proof rows are visible before the transcript

    Observed
    The public persona page shows a five-row T3 table before asking the reader to accept the broader claim.
    Evidence
    The T3 table lists each lane, the cold-start failure mode, the corpus-backed signal, source class, and buyer decision in one visual review surface.
    Supports
    Supports the requirement that the proof case lead with reviewable rows instead of only narrative or JSON.
  5. 05 Publicly inspectable

    Cold baseline review board is public

    Observed
    The public persona page now shows the cold artifact beside the KBPD-backed artifact and lists five baseline checks before the T3 table.
    Evidence
    The `Cold baseline artifact` section includes role boundary, source standard, claim pressure, public proof path, and language gate checks, each with an evidence-row link.
    Supports
    Supports the requirement that the cold-start contrast be visible as an artifact review surface, not only described in prose.
  6. 06 Publicly inspectable

    Raw screenshots are intentionally withheld

    Observed
    The proof case publishes redrawn mechanisms and explicitly reports zero public source screenshots.
    Evidence
    Metric row `Public source screenshots: 0`; source artifact list marks retained run captures as private evidence; the run gallery labels itself as a public projection.
    Supports
    Supports the public/private boundary while explaining why the public page shows projections instead of raw captures.
  7. 07 Publicly inspectable

    Five-frame sanitized run gallery is public

    Observed
    The public persona page now redraws the retained run as five frames: brief, corpus, lanes, synthesis, and buyer proof path.
    Evidence
    The `Sanitized run gallery` section shows retained signal, public projection, and buyer read for each frame, plus an evidence-row link.
    Supports
    Supports the requirement that raw persona run captures get a publication-safe gallery projection before the case becomes a stronger flagship proof.
  8. 08 Publicly inspectable

    Capture review board maps private screens to public proof

    Observed
    The public persona page shows five capture-review rows that connect retained screenshot classes to public frames, redaction boundaries, and exact evidence-row links.
    Evidence
    The `Capture review board` section includes panel launch, architecture lens, journey/verification lenses, operations lens, and business-language gate rows, each with retained signal, public frame, redaction boundary, and an evidence link.
    Supports
    Supports the requirement that private persona screenshots become a usable public proof path rather than an unreviewable screenshot dump.
  9. 09 Publicly inspectable

    Four corpus source classes are represented

    Observed
    The public mechanism shows role roster, prompt archive patterns, prior decisions, and project source as inputs to the run.
    Evidence
    Metric row lists four corpus sources; source artifact list names role registry and prompt archive sources.
    Supports
    Supports the retained-corpus claim without exposing private source material.
  10. 10 Publicly inspectable

    HERMES and cold-reader gates are named

    Observed
    The public case names the public-language lane and the cold-reader gate as corpus owners before the result becomes buyer-facing copy.
    Evidence
    Corpus ownership cards and the T3 proof-row table identify HERMES, archive discipline, session memory, and cold-reader review as separate responsibilities.
    Supports
    Supports the claim that the prompt corpus is managed as a reusable product asset rather than hidden chat history.

Evidence Flow

Read the proof as a chain, not a claim.

Each case has to survive the same path: starting condition, system-backed change, visible result, retained evidence, stated limitation, and a next action.

Specialist lanes 5
Public source screenshots 0
Corpus sources 4

Council mechanism

The buyer should see the run form before reading the transcript.

Private screenshots become a public mechanism: one hard brief, retained corpus, specialist evidence lanes, one synthesis, and a purchase path that can be inspected locally.

Color-separated proof

Five lenses keep their color until the synthesis resolves.

The point is not that the personas have names. The point is that each lens carries a different source class and leaves a visible mark on the final public artifact.

Open evidence row

Cold baseline artifact

Review the baseline before the council earns trust.

The cold start is not weak because it is badly written. It is weak because a buyer cannot inspect which job produced which judgment.

Cold artifact

One blended recommendation

Architecture, journey, verification, operations, and business advice arrive in the same voice. The result can be useful, but the review path is missing.

  • No visible lane ownership
  • No source-class trail
  • No proof-strength boundary until a reviewer adds one
KBPD-backed artifact

Five jobs before synthesis

The same brief is routed through specialist lanes and retained corpus. The synthesis can still be challenged, but the challenge has handles.

  • Lane ownership remains visible
  • Source classes are named
  • Public language and limitations have explicit owners
Check Cold artifact risk Council artifact signal Evidence
01 Role boundary The answer sounds coherent, but the roles are fused into one voice. Each lane has a named job before synthesis. Open row 02 Source standard The reader cannot see which source class shaped each judgment. Role guide, prompt archive, prior decisions, and project source are named. Open row 03 Claim pressure The answer can state confident plans before showing what is proven. Verification labels live, claimed, retained, and unproven states. Open row 04 Public proof path The artifact reads like a transcript summary, not a review surface. The T3 table turns the run into rows a buyer can inspect. Open row 05 Language gate Internal method language can leak into buyer-facing copy. HERMES and cold-reader review own the public-language pass. Open row

Evidence row: the public page now carries a dedicated baseline review board, so the cold-start contrast is a visible artifact rather than only a paragraph.

Open baseline evidence row

T3 proof rows

Read the proof as rows, not as a transcript.

The table makes the cold contrast inspectable: what the prompt-only answer tends to collapse, what the corpus-backed lane separates, which source class is involved, and what decision a buyer can make.

Lane Cold baseline Corpus-backed signal Source class Buyer decision Evidence
01 Architecture Hephaestus One plan blends structure, risk, and sequencing. Dependency order is named before synthesis. Role guide + project source Can the buyer trust the build order? Open row 02 Journey Artemis One recommendation hides where the path breaks. The end-to-end thread and breakpoints are separated. Role guide + prior decisions Can the buyer see the customer path? Open row 03 Verification Cassandra Plausible claims arrive without a claim-strength gate. Live, claimed, retained, and unproven states are labeled. Review standard + source memory Can the buyer see what is not proven? Open row 04 Operations Mnemosyne Release reality and handoff state are implied. Session memory and release state become explicit context. Session trail + prompt archive Can the next run continue without restart? Open row 05 Business Argus + HERMES Value proof and buyer language are mixed into generic advice. Audience, value proof, and public-language pass are routed. Prompt archive + cold-reader gate Can the result become a sellable proof? Open row

Sanitized run gallery

The private screenshots are not pasted into the site. This gallery redraws the usable evidence structure: what was retained, what is safe to show, and what a buyer can decide from each frame.

01

Brief arrives

Retained signal
Private prompt and project references
Public projection
One hard planning question, no private text shown
Buyer read
The buyer sees the task shape before judging the result.
02

Corpus attaches

Retained signal
Role guide, prompt patterns, prior decisions
Public projection
Four context classes named without exposing private material
Buyer read
The buyer can tell what context entered the run.
03

Lanes run separately

Retained signal
Parallel lane outputs and background task traces
Public projection
Five lanes shown before synthesis: architecture, journey, verification, operations, business
Buyer read
The buyer can audit which lens found which constraint.
04

Synthesis keeps the trail

Retained signal
Final recommendation with private details removed
Public projection
Decision, risks, limits, and next action remain tied to lane signals
Buyer read
The buyer can accept or reject the claim without reading a transcript.
05

Public proof path

Retained signal
Raw captures stay internal until redaction clears
Public projection
T3 rows, evidence facts, source trail, and local starter-kit inspection
Buyer read
The buyer gets a review path instead of a screenshot dump.

Evidence row: the gallery is a public projection, not raw screenshot evidence.

Open gallery evidence row

Capture review board

Trace each private capture into a public frame.

Raw screenshots can prove a run existed, but they do not make a buyer-safe case. The review board shows what each capture contributed, how it was redacted, and where the public page lets a reviewer inspect the same signal.

Retained signal
role names, background task fanout, source-read sequence
Public frame
council mechanism + corpus source classes
Redaction boundary
workspace paths, file paths, and private account details removed
Open evidence row
Retained signal
dependency order, proof-gate language, release-risk framing
Public frame
Hephaestus lane + architecture T3 row
Redaction boundary
issue keys and unreleased planning detail collapsed to public source classes
Open evidence row
Retained signal
where the customer path breaks and what is live versus claimed
Public frame
Artemis/Cassandra rows + limitation language
Redaction boundary
client board names and operational URLs withheld
Open evidence row
Retained signal
release reality, session continuity, and board-state checks
Public frame
Mnemosyne row + source-memory ownership
Redaction boundary
raw tables rewritten as retained-source summaries
Open evidence row
Retained signal
value routing, ugly-slide pressure, cold-reader review
Public frame
Argus/HERMES row + buyer proof path
Redaction boundary
internal vocabulary rewritten as public purchase and proof language
Open evidence row

Evidence row: the capture board makes the publication boundary explicit before the reader reaches the retained-source register.

Open capture-board evidence row
01

HERMES

public language and launch copy

02

Archive

prompt records and reusable frames

03

Mnemosyne

session trail and continuity

04

Janus / Stranger

cold-reader clarity checks

This is the content strategy: every strong internal run becomes a visual mechanism, then a proof case, then a method passage, then a local inspection path.

See role gallery Inspect the starter kit

Claim

Role-specific context and a retained prompt corpus produce more inspectable work than a single generic assistant because each lane has a different evidence job before synthesis.

The visible difference is auditability. The council run lets a reviewer inspect which role, source class, and corpus pattern found which constraint before accepting the synthesis.

Before / after

Same use case. Different starting conditions.

Before · cold-start recommendation

Single assistant response

A generic assistant can produce a plausible plan from the prompt alone, but the reasoning lanes, source assumptions, and evidence standards are fused together.

One answer mixes architecture, journey, verification, operations, and business concerns without showing which knowledge source shaped each judgment.

After · corpus-backed synthesis

Council fanout

The same hard brief is routed through specialist lenses, prompt archive patterns, and retained project source before one synthesis is written.

Architecture names dependency order. Journey names where the path breaks. Verification names what is live versus claimed. Operations names release reality. Business names the value proof.

Screenshots to public proof

The marketable claim is the review path, not the source transcript.

The private captures show how specialist lanes come into being. The public proof redraws that sequence as a sanitized product story: one hard brief, retained corpus, separate evidence jobs, visible synthesis, and a buyer path into local inspection.

Purchase path

Turn the proof into a local inspection path.

Local MCPInspect the local access path.WhitepaperRead the public method.Starter kitSee license and purchase boundary.
Compare the role samples See the starter kit

Visual evidence

See the mechanism, not just the claim.

corpus trace

Cold start vs. corpus-backed council

Sanitized projection
The public claim is not that a name improves prose. The claim is that retained corpus and role-specific prompt archive patterns make the output easier to audit.

transcript storyboard

Sanitized transcript storyboard

Sanitized projection
The screenshot evidence stays internal. The public artifact redraws the mechanism: role guide, prompt archive, source memory, task fanout, lane reports, synthesis.

Corpus ownership

The prompt corpus has owners.

KBPD does not treat prompts as disposable text. Public copy, archive discipline, session memory, and cold-reader review each have a named lane before the result becomes a proof or purchase surface.

01

Hermes

public-language patterns

Turns the council result into buyer-readable copy, proof narratives, whitepaper passages, and channel-specific launch surfaces.
02

Archive / Hub Librarian

prompt archive discipline

Preserves role instructions, prompt patterns, reusable briefs, and source references as inspectable corpus assets.
03

Mnemosyne

operational memory

Keeps the session trail, prior choices, limitations, and handoff state available before the next council run begins.
04

Janus / Stranger

buyer-legibility gate

Reads the synthesis like a first-encounter buyer and removes vocabulary that only makes sense inside the project.

Evidence

What this case actually carries.

Run metadata

Date
June 2026
Sample
One sanitized council run
Provenance
Public projection of a retained multi-lane planning run
Reproducibility
The public projection is reproducible as a proof pattern; the retained source requires redaction before publication.

Metrics

  • 5 Specialist lanes Architecture, journey, verification, operations, and business lenses are shown separately.
  • 0 Public source screenshots The public case is redrawn from sanitized structure; private screenshots remain internal.
  • 4 Corpus sources Role roster, prompt archive, prior decisions, and project source are represented in the public mechanism.
  • demo Claim boundary This proves the pattern is inspectable, not that it wins every planning task.
  • 5 T3 proof rows Five public rows connect cold baseline, specialist lane, evidence signal, source class, and buyer decision.
  • 5 Baseline checks Five public checks make the cold-start artifact reviewable before the reader reaches the T3 table.
  • 5 Public run projections Five redrawn gallery frames show the retained run structure without publishing raw screenshots.
  • 5 Capture review rows Five publication-safe rows map private capture classes to public frames, redaction boundaries, and evidence links.
  • 5 Color-separated lenses Five persona lanes now render as layered color-separation plates before the longer mechanism and transcript surfaces.

Source artifacts

  • Council projection sanitized diagram · public projection
  • Raw run captures internal evidence · internal source retained
  • Persona roster role registry · internal source retained
  • Prompt archive corpus source · internal source retained
  • Prompt archive patterns corpus source · internal source retained
  • T3 proof-row table sanitized review table · public projection
  • Cold baseline review board sanitized review table · public projection
  • Sanitized run gallery redrawn run projection · public projection
  • Capture review board redacted capture map · public projection
  • Color-separation lane board sanitized visual proof surface · public projection

Limitations

What this case does not prove.

Next

Follow this proof into the product.

Compare the role samples Talk to us