PikaFam Project Room

Access code required

This page has a lightweight access gate so it does not get opened casually from the projects list.

*** BEST VIEWED WITH CURIOSITY ENABLED ***
Centering Intelligence
TCGMarketPulse + PulseLine
Mini-site
Overview 01 Implementation 02 OpenClaw instance 03 Skills 04 Architecture 05 Roadmap 06
Scope: implementation planning for centering eligibility. This does not claim final grading outcomes.
NemoLink / Projects / TCG Centering / Delivery plan
Draft spec updated Aug 20, 2026

Delivery plan

Roadmap and quality gates

Phased implementation plan with beta definition of done, operational risks, and launch quality gates.

6 phasesBeta gatesProduction risks

Implementation roadmap

Build in phases so each layer can be validated.

PhaseBuildOpenClaw workExit criteria
0 — RulebookPSA/CGC/BGS profiles and SJC/SGC placeholderCreate rules skill and JSON validatorRules load in tests with effective dates and source notes
1 — Local worker MVPQuality, contour, homography, manual handles, ratio mathCreate vision skill and command wrapperClean golden-set images measure within ±1–2%
2 — PulseLine prototypePhoto-in, report-out chat workflowCreate intake and report skillsDiscord photo returns result or retake guidance consistently
3 — TCGMarketPulse integrationAPI job records, artifact links, card-detail panelAdd config and integration wrapperAnalysis attaches to a card/collection record
4 — Layout intelligenceClassifier and segmentation-assisted boundary detectionAdd model profile config and eval reportsManual adjustment rate drops on modern and full-art cards
5 — Calibration loopReturned grade capture and model/rules evaluationCreate admin skill and scheduled eval summaryAccuracy tracked by card type, grader, and confidence band

Definition of done for beta

  • Front/back image analysis works from OpenClaw and product upload.
  • Every result has overlay, confidence, uncertainty, and warnings.
  • Low-quality images do not produce confident grading language.
  • Rules engine supports axis-specific BGS logic.
  • Manual corrections are stored for model improvement.

Risks to resolve before production

  • Exact SJC grader identity and standards.
  • Full-art and borderless measurement semantics.
  • Retention policy for user-submitted card images.
  • Whether slabbed/toploader images are supported at launch.
  • How returned grades are linked back to analysis records.
©2026 - NemoLink by Nemo Prime - Centering intelligence brief