Library / Sections / Research

Research

Foundry discovers the evidence universe. SISO Knowledge preserves it. Evidence Engines turn it into traceable claims. Frontier Questions keep the highest-leverage questions and their answer lineage alive.

R—01
01
SISO Knowledgeknowledge_system · active
02
SISO Foundryapplication · active
03
SISO Evidence Enginesbundle · active
04
SISO People Graphknowledge_system · experimental
05
SISO Book Libraryknowledge_system · experimental
06
UNFUCK — whole-project ownership promptplaybook · active
07
SISO Component Bankknowledge_system · active
08
SISO Repo Bankknowledge_system · active

God Questions Observatory. 4 partial · 1 researching · 3 scoped across 8 standing questions.

Program substrate. 3 questions · 7 assumptions · 10 evidence connections · 8 action/learning links.

Questions drive the portfolio. Research state is distinct from public answer maturity; artifact-free metadata seed Releases are not accepted answers.

Frontier Questions · God Questions

Questions worth answering again.

8 standing questions with stable identities, explicit evidence scopes, and versioned-answer contracts. Metadata seed Releases are not accepted answers.

GQ-001partial

How should the complete agent workspace be designed across code layout, agent architecture, memory, enforcement, and agent-generated user interfaces?

How to lay out a new agent workspace — what to enforce mechanically, what to leave to convention, and which patterns to refuse to repeat.

Steward: SISO Open Source Foundation · Lifecycle: active · Freshness: review due

Next useful work: Run one cold reconstruction task against the typed workspace boundary and compare errors, intervention, and elapsed time with the prior evidence path.

Public answer: not released

hybrid
GQ-002partial

What changes would make the complete provider-neutral SISO agent layer ten times more effective per unit of time, attention, and compute?

Where to spend the next unit of agent-layer engineering effort: fixing an existing path's economics versus substituting a model or harness.

Steward: SISO Open Source Foundation · Lifecycle: active · Freshness: review due

Next useful work: Run one paired agent task with and without the evidence-aware context and receipt path, including retrieval and repair overhead in the comparison.

Public answer: not released

hybrid
GQ-006partial

How should the information organ continuously gather, preserve, query, and push valuable evidence into authorized agent workflows and self-improvement loops?

Whether to trust a published Library number without re-deriving it, and whether an evidence path is sound enough to attach to a standing question.

Steward: SISO Open Source Foundation · Lifecycle: active · Freshness: not declared

Next useful work: Define the minimum public-safe question program before scheduling research.

Public answer: not released

hybrid
GQ-008partial

Which available model should own each recurring work shape after quality, reliability, tool use, latency, quota, and cost are evaluated together?

Which model owns each recurring work shape, and whether an observed cost gap justifies changing the model or fixing the gateway path it was measured through.

Steward: SISO Open Source Foundation · Lifecycle: active · Freshness: not declared

Next useful work: Define the minimum public-safe question program before scheduling research.

Public answer: not released

hybrid
GQ-009researching

How should the Great Library make God Questions visible, composable, continuously updated, and connected to evidence, decisions, approved action, and capability learning without becoming the corpus warehouse or execution runtime?

Which registry fields, Library projections, evidence interfaces, and action-return contracts should become the standard infrastructure for every SISO God Question?

Steward: SISO Open Source Foundation · Lifecycle: active · Freshness: review due

Next useful work: Stand up the operating loop on always-on hardware: a single agent whose domain is improving the Library/Foundry/people-graph system, with a public dossier surface showing data bucket counts, question states, and what it changed and when. Evidence selection must be question-driven given the measured extraction ceiling.

Public answer: not released

hybrid
GQ-010scoped

How do we increase the People Graph's research and decision value by 100× without treating row count as value or weakening identity quality, provenance, rights, privacy, reproducibility, and reversibility?

Choose which source classes, identity methods, temporal and claim layers, query surfaces, and operating loops to fund, merge, revise, or kill for the next evidence-backed People Graph release.

Steward: SISO Open Source Foundation · Lifecycle: active · Freshness: not declared

Next useful work: Define the minimum public-safe question program before scheduling research.

Public answer: not released

hybrid

Research systems

The machinery behind the answers.

8 independently owned Works.

knowledge_systemactive

SISO Knowledge

The durable knowledge-production system for acquiring, curating, storing, indexing, relating, and retrieving SISO knowledge while keeping large or rights-bearing data outside Git.

gls:work:44d2c4da-9998-4d60-b6d1-93465ac25302
applicationactive

SISO Foundry

A source-mining and reuse-knowledge platform for discovering, evaluating, preserving, and navigating valuable external and SISO source.

gls:work:ec664d93-df93-48c5-be40-5d0165886c01
bundleactive

SISO Evidence Engines

A Research package family for source-grounded knowledge items, credibility-ranked engineering principles, and evidence-gated improvement ideas.

gls:work:231aab36-e763-4374-bc94-3a09fd3bf49f
knowledge_systemexperimental

SISO People Graph

An independently addressable Research Work for durable person, identity, work, topic, and temporal graph contracts, with mutable databases and source snapshots kept outside the Great Library registry.

gls:work:2b03baa8-6dab-4e91-ab4c-9b8db4543236
knowledge_systemexperimental

SISO Book Library

An independently addressable Research Work for lossless public-domain book metadata, people and role exports, subject and classification edges, text-quality gates, and content-location contracts.

gls:work:35d3f881-43f3-4da5-87b0-abf694f387ef
playbookactive

UNFUCK — whole-project ownership prompt

A reusable prompt that gives one frontier agent accountable ownership of a whole project, paired with a one-page dispatch brief, the ranked God Questions with written falsifiers, eight machine-readable research contracts, and eleven agent skills for search, verification, classification and delegation.

gls:work:a5ba6f03-f314-46c5-b106-4d4587446dbb
knowledge_systemactive

SISO Component Bank

An offline, queryable index of 8,538 21st.dev UI components: per-component metadata, 7,975 previews, real TSX source for 6,212 (5,210 from the CDN pass, 3,507 from the shadcn-registry pass with classification axes), a hand-curated layer with verbatim human feedback, and the harvester and query scripts that built it.

gls:work:ce031834-4149-42af-8d86-d9a05398fa24
knowledge_systemactive

SISO Repo Bank

The derived, queryable GitHub reuse banks exported read-only from the Foundry identity database: 23,778 rated repositories with category, capability tag and licence, 126 best-per-capability picks, 70 contract cards, adoption-versus-fame for 2,722 packages, the full 226,968-row placement table, a 56,674-repository star farm, and a scrubbed 3,010-row capability shelf.

gls:work:538c4491-8045-4526-9bcd-0268887601d0