Log — H3 Measurement KB

Append-only record of operations on this KB. Entries start with ## [YYYY-MM-DD] <op> | <summary> for easy grep/parsing.

Operations: init, ingest, query, lint, discover, schema-update.

[2026-05-20] init | KB scaffold created

[2026-05-20] ingest | LearnerStudio MMI brief → 3 concepts, 11 entities

User-supplied PDF (LS-mmi-h3.pdf, 8 pages) — no canonical URL; provenance recorded on the source page. Foundational ingest for the KB.

3 concept pages: third-horizon, five-level-framework, implementation-gap.
11 entity pages: LearnerStudio, The Study Group, EdInstruments, Mathematica, Urban Institute, Learning Engineering Tools Competition, Search Institute, Transcend, Learner-Centered Collaborative, ETS, Lemnis.
Asset extraction deferred: pyramid graphic from PDF page 3 → planned wiki/assets/learnerstudio-mmi-h3-2025/five-level-framework.png. No macOS PDF→image tooling available without adding a Python dependency; backfilling separately.
Promotion bar applied: Big Thought, the History Co:Lab, and iThrive Games are named once each as SXSW facilitators — kept inline on the source page rather than promoted. Carnegie Foundation is a co-mention with ETS; not enough material to stand alone. Babak Mostaghimi and Jason Weeby are listed as authors but have no substantive positions of their own in the source — kept as inline mentions on learnerstudio.md. Shereen El Mallah, Brookings, OECD, Aurora Institute, Margulf Foundation, and the other ~40 participant-roster orgs are noted but not promoted on this pass.

[2026-05-20] schema-update | Initial schema authored alongside ingest

schema.md drafted as part of the bootstrap, capturing the research question, high-signal criteria, places worth looking, and a 10-item open-questions watchlist seeded from the MMI tensions section and the demo questions in the strategy doc. Not a revision — first authoring.

[2026-05-20] discover | Seeded second-pass; 5 candidates surfaced, 3 approved

Web search for external (non-LearnerStudio canon) sources substantiating the MMI brief's claims about existing actors. Candidates surfaced:

EdInstruments — About page (https://edinstruments.org/about). Approved.
Education-to-Workforce Indicator Framework — Mathematica + Gates (https://educationtoworkforce.org/). Approved.
Search Institute — Developmental Relationships Framework (https://searchinstitute.org/resources-hub/developmental-relationships-framework). Approved.
Aurora Institute — State of CBE Policy Map. Held; net-new entity, queued for a future ingest where it can establish Ecosystem-level coverage.
Measuring Forward Network — Clearing the Path for Assessment Innovation. Held; freshest material in the candidate set, but a net-new entity and adjacent to the core MMI conversation.

[2026-05-20] ingest | 3 URL sources → 3 stub entities substantiated

New sources: edinstruments-about, education-to-workforce-framework, search-institute-developmental-relationships-framework.
Substantiated stubs: EdInstruments, Mathematica, Search Institute. Each was previously a one-paragraph stub from the MMI brief's second-hand description; each is now backed by a direct source page.
No new entities created on this pass — by design. Each ingest deepens an existing actor rather than fanning out, which keeps the demo legible.
Cross-source pattern surfaced (worth flagging here rather than burying): the MMI brief slots Mathematica and EdInstruments in the same "cataloging and curation" tier, but the direct sources reveal a sharp distinction. EdInstruments is catalog, not evaluator — by explicit choice. Mathematica's E-W Framework is curated and prescriptive — 99 recommended indicators, equity centered. The brief reads them as similar; they're structurally different responses to the same gap.
Search Institute ingest exposed a soft contradiction: the MMI brief credits Search Institute with "validated survey tools," but the Developmental Relationships Framework page itself doesn't name or link those instruments. Asserted upstream but unverified at the source. Flagged as the highest-priority follow-up on that entity.

[2026-05-20] query | Actors across the five-level framework

Question: where does each surveyed actor sit on the five-level framework, and where are the gaps?
Filed as analysis page. Claim: Individual is crowded with mature work; Relationships rests on Search Institute alone; Unit has only capacity-building orgs (no shipped frameworks); Networks is empty among actors surveyed; Ecosystem has partial coverage from the system conditions tier of Mathematica's E-W Framework. The MMI brief's "73% Individual" finding describes the measures; the matching pattern at the actor level is sparser still at the upper four levels.
Backlinks added from five-level-framework concept, implementation-gap concept, and search-institute entity. Linked from index.md under Analysis.

[2026-05-20] people | First two person entries promoted

Promoted the two co-authors of the MMI brief to their own category: person entity pages: Babak Mostaghimi and Jason Weeby. Both clear the SPEC promotion bar as protagonists of the foundational source.
Backlinked the MMI source page and the LearnerStudio entity page so prior name-mentions now resolve to the person pages.
Added a ## People section to index.md between Entities and Concepts.
Promotion bar: Cassie Crockett and Courtney Garcia (LearnerStudio personnel in the participant roster) are not promoted — named in the roster only, no substantive positions in the brief body. Same for the ~50 other participant-roster names. Re-evaluate as more LearnerStudio canon is ingested.