AGI Jobs v0 (v2) — Demo → AGI Jobs Platform at Kardashev II Scale

Vision: model civilization-scale coordination across energy, compute, work, logistics and human governance. This executable demo explores that vision with deterministic synthetic inputs; it does not demonstrate deployed Kardashev-II infrastructure or measured economic superiority.

This dossier is the living operator manual for demo/AGI-Jobs-Platform-at-Kardashev-II-Scale. It explains how to reproduce the simulation, inspect its assumptions, interpret its evidence and test changes. The command deck is read-only; it has no wallet connection or transaction submission.

Start with a useful question

Can the modeled mission fit its available power, compute capacity, validator coverage and governance constraints? The demo computes a reviewable answer from JSON inputs, then shows the allocations, warnings, stress scenarios, unsigned call proposals and diagrams. It does not contact energy providers, bridge networks, identity services or a blockchain. Historical fixture dates remain intact so that runs are reproducible; they are not claims of fresh observations.

The legacy model remains available and is never an automatic fallback for the canonical orchestrator. Its default output is a cache directory, keeping the canonical checked-in snapshot intact.

Computer work: practical scopes and a path to scale

Every command deck now prominently includes Computer work: ten self-contained synthetic task scopes, exact JSON drafts, required deliverables and independent acceptance criteria, plus an interactive annual capacity planner. Start with one verifiable digital job, then model workers and human review capacity together.

Read the complete computer-work runbook for the actual Chromium Supplier Desk fixture, fault injection, OpenClaw Codex Computer Use readiness, ChatGPT Work permissions, operator admission, evidence, reconciliation and contract finalization. The downloadable drafts use the repository's current worker schema; they grant no authority and contact no provider. The planner is a separate local exercise, not an input to the civilization ledgers.

USD 40 trillion/year is a user-supplied market planning assumption. Neither universal human-task automation nor market capture is established by this demo. Test representative capabilities in a commissioned environment and retain independent review.

🧭 Ultra-deep readiness map

• Location: demo/AGI-Jobs-Platform-at-Kardashev-II-Scale/
• Operating manifest: config/kardashev-ii.manifest.json (council, logistics, verification, and drill cadence).
• Energy & compute telemetry: output/kardashev-energy-feeds.json, output/kardashev-telemetry.json.
• Decision ledger: output/kardashev-orchestration-report.md summarises the last orchestrator pass.
• CI gate: npm run demo:kardashev-ii:ci (enforced on PRs touching this directory).

🚀 Kardashev-II operator quickstart

From the repository root, use the pinned toolchain:

nvm install
nvm use
npm ci
npm run demo:kardashev-ii:orchestrate
npm run demo:kardashev-ii:ci
npm run demo:kardashev-ii:serve

Open http://127.0.0.1:4175/. Stop with Ctrl+C. The server binds only to loopback; it serves a snapshot and performs no wallet or provider actions. Use -- --port 4176 if the port is occupied. For the other variants, generate them with the commands above, then serve with npm run demo:kardashev-ii:serve -- --profile stellar-civilization-lattice or --profile k2-stellar-demo. --generate explicitly regenerates the selected profile before serving; a failed run stops startup instead of changing models.

Expected main result: three programmes, eleven stress scenarios, no critical scenario, and five scenario warnings in the bundled baseline. The equilibrium warning is intentionally visible; “generated” does not mean every indicator is nominal. Inspect output/kardashev-run-manifest.json for input SHA-256 hashes and simulation scope. Inspect kardashev-stability-ledger.json for individual checks, not just its aggregate score.

What check mode does: --check / --ci recomputes artifacts, compares them byte-for-byte, checks dashboard assets, and fails on malformed inputs or critical/high model-readiness failures. It writes nothing, including when output is missing. Medium/low advisories remain visible. --reflect prints additional diagnostics; --check is not a generation command.

A controlled experiment

mkdir -p /tmp/k2-experiment
cp -R demo/AGI-Jobs-Platform-at-Kardashev-II-Scale/config /tmp/k2-experiment/config
npm run demo:kardashev-ii:orchestrate -- --config-root /tmp/k2-experiment --output-dir /tmp/k2-result
npm run demo:kardashev-ii:serve -- --output-dir /tmp/k2-result

Change one input in the copied manifest, such as a bridge latency beyond its failsafe limit, and rerun. The generated ledgers show the failed checks and the command exits nonzero. Restore the input and confirm recovery. Duplicate IDs, missing references, cyclic dependencies, malformed timestamps, non-finite numbers and inconsistent thermostat bounds fail validation before generation. Paths resolve from the current working directory; misspelled flags and missing values fail instead of silently selecting defaults.

How to read the model

• Power and storage: GW is power; GWh is energy. Regional allocation must fit captured capacity after its reserve margin and the thermostat budget. Planned Dyson capacity can exceed regional load; that surplus is not reconciliation drift. The retained delta fields show distance between those assumptions, not disagreement between independent measurements.
• Model scores: dominance, equilibrium, welfare, confidence and “unstoppable” are named heuristics derived from the fixture. They are not empirical probabilities, AGI capability benchmarks, security certifications or financial forecasts.
• Thermostat: legacy fields named minKelvin, targetKelvin and maxKelvin are normalized control parameters, not measured stellar temperatures.
• Governance: the Safe JSON is an unsigned proposal containing placeholder addresses. Selector decoding and hashes establish structural consistency only. Do not submit the bundled batch to a wallet or production chain.
• Boundary: authentic signing, independently verified provider integration, security review and target-network commissioning remain separate steps. The demo's scale is an architectural scenario, not a physical Kardashev classification measurement.

Troubleshooting

🧱 Architecture overview

flowchart TD
    Council[Interstellar Council Manifest] --> MissionHub[Kardashev-II Mission Hub]
    MissionHub --> Ledgers[Energy • Settlement • Consistency Ledgers]
    MissionHub --> Dashboards[UI Dashboards]
    MissionHub --> CI[npm run demo:kardashev-ii:ci]
    Ledgers --> Governance[Guardian & Owner Review]
    CI --> Governance
    Dashboards --> Operators((Mission Owners))

• scripts/run-kardashev-demo.ts ingests the manifest and fabric definitions to regenerate ledgers under output/.
• Dashboards in index.html + ui/dashboard.js ingest these ledgers to project readiness metrics for mission owners.
• CI validation (scripts/ci-validate.ts) replays orchestrator checks and enforces documentation parity.

🪪 Identity lattice & trust fabric

• Defined inside config/kardashev-ii.manifest.json.identityProtocols for global and federation-specific anchors.
• Identity anchors, attestation latency and coverage assumptions appear in output/kardashev-telemetry.json; owner-proof JSON instead describes encoded governance calls.
• Actual guardian approvals are external to this simulation. No signature is collected or authenticated by this runner.

🛰️ Compute fabric hierarchy

flowchart LR
    Earth[Solara Earth Core] --> Mars[Solara Mars Relay]
    Mars --> Orbital[Orbital Research Array]
    Orbital --> DeepSpace[Deep Space Forge]
    DeepSpace --> MissionHub

• Details live under config/kardashev-ii.manifest.json.computeFabrics.
• Availability, failover partner, and energy draw metrics synchronise into output/kardashev-telemetry.json.
• output/kardashev-mermaid.mmd auto-renders the hierarchy for downstream dashboards and is loaded by ui/dashboard.js.

🔌 Energy & compute governance

• Energy parameters sourced from config/energy-feeds.json and config/kardashev-ii.manifest.json.energyProtocols.
• Governance playbook stored in output/governance-playbook.md with explicit guardian cadence.
• Thermostat ranges are inputs to local model calculations. This runner does not send them to an external thermostat service.

⚡ Live energy feed reconciliation

• output/kardashev-energy-feeds.json captures regional supply; output/kardashev-energy-schedule.json cross-verifies dispatch windows.
• scripts/run-kardashev-demo.ts performs kahan- and pairwise-sum comparisons to eliminate reconciliation drift.
• The configured feed and cross-verification tolerances determine warnings; inspect the emitted tolerances rather than assuming a fixed ±0.1% rule.

🔋 Energy window scheduler & coverage ledger

• Scheduler logic resides in scripts/run-kardashev-demo.ts (energy schedule calculation) and writes to output/kardashev-energy-schedule.json.
• Coverage and reliability projections appear in output/kardashev-energy-schedule.json; fabric-ledger JSON describes shard routing and coverage references.
• Keep signed operational approvals separately; the generated owner-proof file is overwritten reproducibly and contains no authenticated signatures.

🚚 Interstellar logistics lattice

• Logistics corridors declared in config/kardashev-ii.manifest.json.logisticsCorridors.
• Runtime health published to output/kardashev-logistics-ledger.json with capacity, jitter, and buffer-day metrics.
• Logistics visualisations refresh in index.html via the renderLogistics handler inside ui/dashboard.js.

🕸️ Sharded job fabric & routing ledger

• Federation shards and job registries defined in config/fabric.json.
• Routing results captured in output/kardashev-task-ledger.json, mapping tasks to shards and guardians.
• Cross-reference validation checks declared identifiers; it cannot guarantee operational routing or consensus.

🎛️ Mission directives & verification dashboards

• Owner directives under config/kardashev-ii.manifest.json.missionDirectives map to Safe transaction bundles.
• Verification dashboards consume generated JSON ledgers. Read output/kardashev-orchestration-report.md alongside them; output/kardashev-report.md belongs to the legacy runner.
• UI entry point: index.html with components rendered by ui/dashboard.js.

🌐 Settlement lattice & forex fabric

• Settlement exposures and forex references export to output/kardashev-settlement-ledger.json.
• Treasury data originates from config/kardashev-ii.manifest.json.interstellarCouncil addresses.
• These are modeled exposures and finality assumptions. The runner releases no payments; the legacy kardashev-report.md is not a transaction receipt.

♾️ Consistency ledger & multi-angle verification

• output/kardashev-consistency-ledger.json records agreement between local calculations. Input fingerprints are in the run manifest; no guardian signatures are produced.
• CI recomputes deterministic artifacts and readiness checks. Keccak256 hashes support consistency checks, not distributed consensus or signer authentication.
• Diff noise is surfaced in kardashev-orchestration-report.md under the “Consistency” section.

🔭 Scenario stress sweep

• Stress vectors embedded within config/kardashev-ii.manifest.json.verificationProtocols (energy models, latency tolerances).
• Full sweep results land in output/kardashev-scenario-sweep.json and are summarized in output/kardashev-orchestration-report.md.
• Schedule a sweep post-change with npm run demo:kardashev-ii:orchestrate -- --reflect to attach introspection notes.

🪐 Mission lattice & task hierarchy

• Hierarchical missions live in config/task-lattice.json and include timelines, autonomy rates, and fallback plans.
• output/kardashev-task-hierarchy.mmd renders the mission tree for rapid situational awareness.
• Guardians cross-link tasks to sentinel coverage inside output/kardashev-task-ledger.json.

🧬 Stability ledger & unstoppable consensus

• System resilience metrics recorded in output/kardashev-stability-ledger.json.
• Thermostat guardrails and pause levers surfaced in output/kardashev-owner-proof.json for council audits.
• CI checks encoded pause/resume proposals and model readiness; it does not replay transactions on a chain.

🛡️ Governance and safety levers

• Pause, upgrade, and deployment levers defined in config/kardashev-ii.manifest.json.missionDirectives.ownerPowers.
• Guardian drill cadence (hours/minutes) is a configured assumption; see missionDirectives.drills in the manifest. The runner does not execute drills.
• Align with repo-level emergency playbooks under demo/agi-governance/ for multi-mission escalations.

🗝️ Owner override proof deck

• Owner-call structure and payload hashes are recorded in output/kardashev-owner-proof.json; the unsigned batch is in kardashev-safe-transaction-batch.json.
• The orchestration report summarizes structural checks, not approvals from actual signers.
• No scheduled compliance export is performed by this demo. Archive reviewed artifacts separately when required.

📦 Artefacts in this directory

• config/ — manifest, fabric topology, energy feeds, and mission lattice JSON.
• scripts/ — TypeScript automation for orchestration and CI enforcement.
• output/ — generated ledgers, complete offline dashboards, and Mermaid sources.
• output/kardashev-equilibrium-ledger.json — equilibrium scoring across energy, allocation, welfare, logistics, and compute fabrics.
• ui/ — static dashboards consuming the latest artefacts.
• k2-stellar-demo/, stellar-civilization-lattice/ — specialised sub-demos with their own manifests and CI rituals.
• run-demo.cjs, index.html — launchers for the operator experience.

🧪 Verification rituals

• Per-change: npm run demo:kardashev-ii:ci (required; fails if documentation or ledgers drift).
• Pre-launch: npm run demo:kardashev-ii:orchestrate -- --check to dry-run invariants against new manifests.
• Full publish: npm run demo:kardashev to write refreshed artefacts and dashboards.
• Cross-demo: npm run demo:kardashev-ii-stellar:ci to ensure subordinate lattice states remain aligned.

🧠 Reflective checklist for owners

• Have the input hashes and explicitly simulated evidence scope been inspected?
• Are energy windows (output/kardashev-energy-schedule.json) covering ≥ 1.1× projected demand?
• Do logistics buffers in output/kardashev-logistics-ledger.json exceed the minimums in the manifest?
• Has npm run demo:kardashev-ii:ci produced a ✔ result after your changes?
• Are mission directives in config/kardashev-ii.manifest.json mirrored in the directive cards rendered by ui/dashboard.js?
• Have unsigned proposal checks been kept separate from any real guardian approvals?

Continuous alignment: rerun npm run demo:kardashev-ii:ci after every change in this tree. A passing run establishes the checked model behavior and reproducibility; it does not guarantee real-world consensus.