LGD shelf

The seven flagships 七件旗舰

Each asset is a door. Every page carries the hash, the score and the command to recompute both. Seven downloadable packs, 295 KB altogether, each with its own sha256 in /dist/MANIFEST.json.

assetlicencescoretier
skill-preflight
Zero-token pre-listing check for any SKILL.md against a weighted D1-D10 rubric: trigger design, metadata, structure, ver
MIT0.936A
publish-quality-gate
Four-layer sensitive-data scan (company / machine / personal / confidential) plus falsifiable-claim checks before sharin
CC-BY-4.00.887A
evidence-chain-suite
One-stop evidence chain toolkit: build, check, traceability audit and audit-trail custody in a single skill. Use wheneve
MIT0.970A
medxpert-reg-hub
All-in-one knowledge hub for medical-device registration engineers: global pathways and classification (China NMPA, US F
MIT0.965A
uibc-challenge
UIBC challenge: rules, scoring, and participation entry for evaluating AI agents against user-intent benchmarks, with pu
CC-BY-4.00.970A
skill-doctor
Audits installed skills and tells you which actually work: usable, broken, empty shell or duplicate, ranked by grade; de
MIT0.964A
lgd-gate-methodology
LGD gate methodology: a from-zero playbook for building open-source trust and safety tooling with registration, evidence
MIT0.970A

/skills.json · /EVAL.md · /ANCHORS.md · /dist/MANIFEST.json · entry page