Lines
Saved output line count.
Outputs inspected
| Variant | Output | Lines | Words |
|---|---|---|---|
| [C1] Manual brief | 00-control-manual-evidence-decision-workspace.md | 9 | 55 |
| [C2] Matched reusable-template control | 01-control-template-evidence-decision-workspace.md | 152 | 1,092 |
| [T] WeaveMark treatment | 02-treatment-promplet-evidence-decision-workspace.md | 1,675 | 6,632 |
Metric definitions and scoring legendOpen for exact meanings.
Words
Saved compiled output word count.
Verbatim snippets
Verbatim source/output material quoted from saved artifacts.
[C1] Manual briefOpen output
Design a local-first web application that helps me collect documents, notes, links, and news; audit claims and evidence; compare explanations and options; make decisions; assign follow-up actions; and preserve provenance.
[C2] Matched reusable-template controlOpen output
- This is the coherent implementation-ready specification for Evidence-to-Decision Workspace. - Treat this specification as the source of truth for a programming agent or human engineer. - Include first-build scope, out-of-scope items, architecture, domain model, durable records, workflows, UI surfaces, automation rules, validation plan,
[T] WeaveMark treatmentOpen output
Build **Evidence-to-Decision Workspace**: a local-first web application where messy sources become traceable claims, evidence, contradictions, explanations, decisions, actions, alerts, and reusable outputs. The product is an evidence command center for:
[T] WeaveMark treatmentOpen source seam
Design @{app_name}: a local-first workspace where messy sources become traceable claims, evidence, contradictions, explanations, decisions, actions, alerts, and reusable outputs.
Contrastive gain/loss scores
Scores compare [T] WeaveMark treatment against [C2] Matched reusable-template control on the -3..+3 scale.
Primary scores are blind* using hybrid-derived-metrics-and-masked-review: anonymous absolute 1..7 scores were frozen before reveal, then converted to the -3..+3 treatment-control scale. Hybrid blind* scoring uses derived metrics for mechanical criteria and masked source/output review for criteria that require actual reading. The masked review is less blind because domain content, source syntax, or style can leak, but this is necessary to avoid replacing readability and integration judgments with weak length/density proxies.
| Criterion | Blind* score | Evidence |
|---|---|---|
| Authoring leverage | +2 | derived-evidence method. Blind* absolute scores: 7 for [T] versus 4 for the strongest control. |
| Information yield | +2 | derived-evidence method. Blind* absolute scores: 7 for [T] versus 4 for the strongest control. |
| Grounded expressiveness | +2 | masked-source-output review method. Blind* absolute scores: 7 for [T] versus 4 for the strongest control. |
| Input readability | +1 | masked-source review method. Blind* absolute scores: 5 for [T] versus 4 for the strongest control. |
| Output readability | +1 | masked-output review method. Blind* absolute scores: 6 for [T] versus 5 for the strongest control. |
| Constraint integration | +2 | masked-source-output review method. Blind* absolute scores: 7 for [T] versus 4 for the strongest control. |
| Reusable abstraction quality | +1 | masked-source review method. Blind* absolute scores: 6 for [T] versus 5 for the strongest control. |
| Total | +11 | Net contrastive gain/loss. |
Metric definitions and scoring legendOpen for exact meanings.
Contrastive score
A -3..+3 judgment comparing [T] against the strongest listed control for each criterion.
Total score
The sum of the seven contrastive criterion scores for one study.
Score color
Green means [T] is better, red means worse, amber means similar; intensity follows magnitude.
What improved and what failed
What improved
- [T] WeaveMark treatment wins source-only leverage: 26.63 versus 16.3 for [C2] Matched reusable-template control.
- [T] WeaveMark treatment wins discounted fact units: 396.5 versus 96.5 for [C2] Matched reusable-template control.
- [T] WeaveMark treatment wins information yield: 1,592.4 versus 1,440.3 for [C2] Matched reusable-template control.
- The treatment converts evidence quality, ACH, contradictions, decision gates, actions, storage, UI, APIs, and AI safety into one architecture.
- It wins leverage and information yield against the matched reusable-template control.
- The final specification has the largest fact-unit count in the study corpus.
What failed or did not improve
- [T] WeaveMark treatment loses information density: 59.8 versus 88.4 for [C2] Matched reusable-template control.
- [T] WeaveMark treatment is much longer: 6,632 words versus 1,092 for [C2] Matched reusable-template control.
- The matched template remains denser.
- Output readability is mixed because the treatment is much longer and requires careful navigation.
- No analyst-task or implementation outcome has been measured.
Interpretation
The strongest realistic application result on total semantic content and yield, though not on compactness. The qualitative claim should include both sides: WeaveMark improves semantic integration where shown, but the measured failures and caveats are part of the result.