Agents
A curated Library section for complete agent stacks, their independently addressable components, and the evidence showing how they work together.
gls:work:41b4544a-2d5b-4f63-92a2-3f25f64e29d0Public registry · Permanent source map
The Great Library gives every cataloged Work a stable public detail page—even when its source repository has no site of its own.
Works stay stable. Repositories and upstream docs remain optional, replaceable locators.
Relationships carry structure. Hierarchy is a versioned view, never hidden in identity.
New here?
The onboarding page gives humans and agents one cold-start route through identity, evidence, releases, snapshots, contribution, and safety.
Versioned projections
Sections are browsable views over stable Works. Moving a Work between sections never changes its identity or source history.
Live index
25 accepted records in this build.
A curated Library section for complete agent stacks, their independently addressable components, and the evidence showing how they work together.
gls:work:41b4544a-2d5b-4f63-92a2-3f25f64e29d0A standing research question about the complete code, agent, memory, enforcement, and user-interface architecture of an agent-native workspace.
gls:work:2ac84615-3f9d-4040-95d6-9d9f39e4dc84A standing research question about making a provider-neutral multi-agent operating layer materially more effective, efficient, observable, and self-improving.
gls:work:e38b054f-b926-4410-b84d-4fd3249fd4f4A parametric standing question for identifying the best implementation of a reusable software primitive across the corpus with quality, rights, and composability evidence.
gls:work:85964f1c-2efe-48b0-8f71-dc32bb668c0cA standing research question about which technologies, categories, techniques, and projects are accelerating and what those changes predict.
gls:work:7782f923-d44e-4a47-8ce6-134e31816673A standing research and systems question about making broad, continuously refreshed evidence reliably accessible to every authorized agent workflow.
gls:work:2144e900-4886-461e-92f1-49571f2e9f74A standing research question about which available model should own each work shape after capability, reliability, latency, quota, and cost are considered together.
gls:work:3b4c4fa6-31f2-4cfc-9542-00f32e200621A standing meta-question about how the Great Library should make SISO's highest-value questions visible, composable, continuously revisable, and connected to evidence and approved action.
gls:work:3e220d53-cd95-4e28-922e-5018e808ee92An externally owned terminal-native multiplexer providing persistent workspaces, tabs, panes, and addressable agent sessions.
gls:work:831c3bc6-36a0-4f88-93a8-55d3240849ccA curated Library section for source discovery, evidence pipelines, knowledge engines, datasets, and durable research artifacts.
gls:work:2fa08dac-937a-44d3-a547-f16ac28c4e25The retired mixed-source repository from which SISO Runtime, capabilities, intelligence, integrations, and research Works are being read, preserved, and reallocated.
gls:work:ee26c795-ac31-4f31-bb8f-e54dbc515baeThe shared state service for task lifecycle and atomic queues, versioned task artifacts, memory, timeline, fleet presence, costs, human questions, delegation, control flags, and governed units.
gls:work:5af5cab4-13d9-48c4-b15c-8c31ba3db10bThe portable lifecycle automation boundary for Claude Code and Codex, with one manifest-driven event map, host-native generated configuration, guarded installation, privacy-reduced state, and adversarial verification.
gls:work:4198f323-9855-43f6-a99a-72143086818eAn experimental package family for bounded agent-runtime adapters with explicit safety, evidence, promotion, and rollback contracts.
gls:work:e88d82a7-3a24-4aaf-8758-ea7372c3797aAn installable operating package of scenario playbooks, prompts, skills, tools, telemetry, and evidence for running coordinated agent work.
gls:work:083503ab-c78e-4e07-ac40-ab9466dcedccThe provider-neutral SISO coding-agent application boundary: explicit configuration, safe host launch, profile composition, task control, checkpoints, bounded context, status, and runtime event contracts.
gls:work:b584bbf7-acde-46a2-b33e-e430ee1df224The one-clone, version-pinned public distribution for the SISO Claude Code and Codex environment, composing independently released skills, hooks, playbooks, agent roles, operating profiles, commands, runtimes, and optional services without merging their repository identities.
gls:work:b9196c61-901f-4105-93d9-899876d17016The intent-holding coordination and boot package that joins a durable identity to current Agent Brain state, then hands work to an explicitly selected agent host.
gls:work:ea2f0de7-95ee-4e62-acde-85375b048719A Research package family for source-grounded knowledge items, credibility-ranked engineering principles, and evidence-gated improvement ideas.
gls:work:231aab36-e763-4374-bc94-3a09fd3bf49fA source-mining and reuse-knowledge platform for discovering, evaluating, preserving, and navigating valuable external and SISO source.
gls:work:ec664d93-df93-48c5-be40-5d0165886c01The 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-b6d1-93465ac25302A project-local operating system and contract bundle for durable agent work, evidence, knowledge, and handoff state.
gls:work:54a52e52-dfbd-4922-8c04-ec02f7fae943A privacy-reduced learning pipeline that turns explicitly supplied Claude and Codex sessions into traceable goals, failures, decisions, intentions, wins, recurring patterns, and durable local recall.
gls:work:3a7863ae-9a72-44bf-86ce-e730bcff3b84The registry and distribution boundary for atomic repeatable capabilities, with evidence-backed classification, bundled installation, isolated skill telemetry, and independent repository promotion only when a skill earns its own lifecycle.
gls:work:8acb11e9-026f-47e3-b620-bb1b969bcbcdThe public human-readable and agent-readable registry for SISO source.
gls:work:9b18cc64-7d5e-4f30-b63d-a0ec30c1f7e1No Works match those filters.
Counts reflect accepted registry files in this build.