Demo · Demonstrator
A team wants to understand why KBPD uses specialist roles instead of one generic assistant voice.
process trace
HERMES case-study frame
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.
A team wants to understand why KBPD uses specialist roles instead of one generic assistant voice.
A general assistant blends strategy, design, critique, and implementation into one plausible but hard-to-audit response.
Each role has a distinct job: decide, explain, verify, translate, remember, build, direct attention, or choreograph state changes.
Samples are static in the first launch.
Proof row
The public role gallery exposes eight launch-safe roles with a role caption, capability statement, and concrete output sample.
Supports the claim that the public product story shows multiple distinct specialist outputs.
Evidence room
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.
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 evidence excerpts
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.
role-gallery-surface-excerpt The public gallery presents eight launch-safe role samples and routes them into decision, copy, verification, and visual/motion surface types.
interactive-boundary-excerpt Static role samples ship first; interactive role calls are reserved for the hosted-service phase.
Claim under review
The proof is separation of concerns. The work becomes easier to review because every output has a named job.
evidence_facts[0] Samples are static in the first launch.
evidence_facts[1] Interactive role calls belong in the hosted-service phase.
evidence_facts[2] Samples are static in the first launch.
Evidence values
Source trail
source-01-role-samples Role samples homepage component · public projection source-02-role-roster Role roster internal source · internal source retained Evidence status mix
Metrics
Source artifacts
Case quality review
This is a cold-review rubric, not a sales claim. It checks whether the case can be understood, inspected, bounded, and acted on.
A reviewer gets the situation, contrast, and buyer decision without reading the JSON.
The case exposes proof rows, values, source handles, and excerpts before the JSON export.
The page names what the case proves and what it does not prove.
The route uses visuals, storyboards, or surface frames to break the proof into inspectable layers.
The proof sends the buyer to an evidence room, artifact link, pricing path, or local run.
Cold-audience clarity: Tighten the opening latch or before/after language when the case still needs explanation.
Open case evidenceEvidence docket
Evidence Flow
Each case has to survive the same path: starting condition, system-backed change, visible result, retained evidence, stated limitation, and a next action.
Claim
The proof is separation of concerns. The work becomes easier to review because every output has a named job.
Before / after
Before · single-lens output
A general assistant blends strategy, design, critique, and implementation into one plausible but hard-to-audit response.
Looks coherent, but the reader cannot tell which role made which trade-off.
After · capability gallery
Each role has a distinct job: decide, explain, verify, translate, remember, build, direct attention, or choreograph state changes.
The site shows concrete samples for the roles a buyer is most likely to understand first, including the visual and motion roles that make Proof Studio more than text.
Role gallery as proof
A buyer should be able to see the difference between a generic answer and KBPD before reading the full method. The gallery now shows how one structured brief becomes separate expert jobs, then public artifacts with claim limits.
The same project source enters every lane, so differences are caused by role and corpus, not by changing the task.
Prompt patterns, prior decisions, role guides, and page intent are explicit before outputs are written.
Architecture, copy, legibility, verification, art direction, motion, memory, and cold-audience review each own a different job.
The outputs land as buyer-visible surfaces: proof frames, whitepaper sections, pricing boundaries, and motion rules.
The claim stays labeled as a demonstrator until richer measured comparisons are ready.
Purchase path
Visual evidence
role gallery
role matrix
surface route
Corpus ownership
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.
Hermes
Archive / Hub Librarian
Mnemosyne
Janus / Stranger
Evidence
Run metadata
Metrics
Source artifacts
Limitations
Next