Work / gls:work:2ac84615-3f9d-4040-95d6-9d9f39e4dc84

GQ-001 · The Agent Workspace

A standing research question about the complete code, agent, memory, enforcement, and user-interface architecture of an agent-native workspace.

Type
research_question
Maturity
active
Section
Research

Research contract · GQ-001

Question

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

Steward

SISO Open Source Foundation

Lifecycle status

active

Research state

partial

Freshness

Current · assessed 2026-08-02 · review by 2026-09-01

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.

Selected release

gls:release:0212a6f4-1860-453b-bb27-2e9ed70a09c4 · program-metadata-1.0.0+ea7f0e4

Public answer release

Not released — the selected Release contains question-program metadata, not an answer artifact.

Evidence mode

hybrid

Source scopes

public agent-workspace repositories · code-layout and architectural enforcement · memory and context systems · agent-facing interface patterns

Answer shape

Evidence-graded principles, reference architectures, rejected patterns, and tests that distinguish universal findings from SISO preferences.

Refresh policy

Re-answer when a major agent platform, workspace architecture, or failed forward prediction changes a load-bearing conclusion.

Publication boundary

public metadata only

Read the God Questions infrastructure constitution →

Assumptions · 2

QA-GQ001-BOUNDARIES · QA · active · medium

A complete agent workspace can integrate its organs through typed interfaces without collapsing their independent ownership boundaries. Scope: agent-workspace architecture · Falsifier: Representative workspace tasks require one organ to own another organ's durable state, evidence adjudication, and runtime authority to remain coherent. · Depends on: none · Evidence: EC-GQ001-RESEARCH-METHOD · Review: 2026-08-02

QA-GQ001-ADDRESSABILITY · QA · active · medium

Stable identities, typed relationships, and explicit receipts reduce the context a cold agent must reconstruct before useful work. Scope: workspace reconstruction · Falsifier: Cold-task comparisons show no meaningful reduction in reconstruction time, errors, or human intervention after adding the addressable contract. · Depends on: QA-GQ001-BOUNDARIES · Evidence: EC-GQ001-LIBRARY-SEED · Review: 2026-08-02

Evidence connections · 2

EC-GQ001-RESEARCH-METHOD · documentation · public

The question-driven research architecture separates identity, discovery, corpus, evidence transformation, and execution responsibilities. Owner: The Great Library of SISO · Supports: QA-GQ001-BOUNDARIES · Challenges: none · Observed: 2026-08-02

EC-GQ001-LIBRARY-SEED · release_receipt · public_metadata_only

The immutable seed exposes publication-safe identity and scope without presenting the partial answer as a public artifact. Owner: The Great Library of SISO · Supports: QA-GQ001-ADDRESSABILITY · Challenges: none · Observed: 2026-08-02

Action and learning lineage · 1

AL-GQ001-DEMAND · epistemic demand · demand only

The current question frame defines learning demand but grants no implementation or budget authority. Owner role: question steward · Status: proposed · Recorded: 2026-08-02T00:34:00+07:00 · Truth: not applicable · Predecessors: none · Assumptions: QA-GQ001-BOUNDARIES, QA-GQ001-ADDRESSABILITY

Relationships

integrates_with

SISO Foundry

depends_on

SISO Knowledge

coordinates_via

SISO Evidence Engines

Provenance

Registry source
registry/works/frontier-question-gq-001.json
Origin
siso
License / redistribution
NONE (not_applicable)

Source & upstream links

Nothing has been declared here yet.