Honest snapshot of Spanda Platform capabilities as of v0.5.0. The Spanda Language (.sd) is
one component; this matrix covers verification, simulation, fleet, packages, and tooling as well.
Use this document to understand what is production-ready, experimental, planned, or deprecated.
Platform overview: platform-overview.md · Release hardening: scope-control.md · Blockers: release-blockers.md
| Label | Meaning |
|---|---|
| Stable | Tested default path; not mock/demo/docs-only |
| Beta | Usable with known limitations; suitable for evaluation |
| Experimental | Works with caveats; optional live backends |
| Preview | Early API; may change without notice |
| Stubbed | Syntax or API exists without full external integration |
| Mock-backed | Default path uses mocks/simulators (not production-ready alone) |
| Planned | Not implemented |
| Deprecated | Replaced; do not use in new programs |
Rules: mock-backed, demo-only, docs-only, untested, and simulated-only features are not Stable. Simulated-only paths must say so explicitly.
Stubbed = syntax or API exists without full external integration.
Broken = known to fail or incomplete in current builds.
| Area | Status |
|---|---|
| Differentiation NOW | spanda demo differentiation, all five NEXT analytics pillars Stable |
| Distributed decisions | Decision trees, offline policy, signed trees, v3 signed traces, persistent escalation, conflict resolution — Stable |
| Enterprise operations | Control Center E1–E4, device pool, gRPC parity — Stable |
| Cognitive & Resilience Architecture | Eleven functional domains — Beta tier |
| Reflex, immunity, homeostasis, attention; fusion, memory, operational memory | |
| Adaptive recovery Experimental; CC Cognitive & Resilience tab Beta | |
architecture · maturity · CI: scripts/cognitive_resilience_smoke.sh |
|
| Solution blueprints | ADAS, Smart Spaces, Spatial Computing — Stable (organizational soak gates separate) |
| Area | Status |
|---|---|
| Native deploy | spanda deploy --target native, compile-native, LLVM IR — Experimental (clang required) |
| ROS 2 interop | spanda ros2 check, rclpy bridge with SPANDA_ROS2_LIVE=1 — Experimental |
| Distributed fleet | fleet orchestrate --remote, agent registry — Experimental |
| CLI install | cargo install --path crates/spanda-cli → binary spanda — Stable |
| Bundled demos | spanda demo without full clone — Stable |
| Plugin system | spanda plugin search|install|enable|disable|list|inspect|trust; spanda.plugin.toml; registry trust tiers; WASM sandbox (wasm-loader); namespaced CLI commands; Control Center /v1/plugins — Experimental |
| Cascading configuration | spanda config, spanda drift, spanda device discover|inspect, spanda network scan, spanda device-tree, spanda map verify; DeviceRegistry + network identity validation; config and agent drift (--baseline, --agent); --config on run/verify/readiness/replay/assurance — Experimental |
| Area | Capabilities |
|---|---|
| Language core | Lexer, parser, AST, type checker, physical units, module/import, structs/enums/traits, match, Result/Option, test blocks |
| AI agents | ai_model, agent, goal, memory, mock-backed LLM/Vision providers by default, ActionProposal → safety.validate() → SafeAction |
| Robotics primitives | robot, sensor, actuator, behavior, task every Nms, state machines, events |
| Hardware profiles | hardware, deploy, requires_hardware, requires_network, SoC/HAL validation |
| Compatibility verification | spanda verify, --target, --all-targets, --simulate, --json |
| Simulation | spanda run / spanda sim, physics-lite 2D backend, lidar/arm/drone models |
| Communication | message, topic, service, action, publish/call/send_goal, in-memory transport |
| Safety validation | Safety zones, max_speed, stop_if, emergency stop, compile-time SafeAction gate |
| Trigger-driven execution | Unified on / every / when / while; event, message, timer, condition, state, safety, hardware, AI, verification, twin |
| Cooperative concurrency | spawn, join, parallel, channels, select, per-task budget { }; TypeScript mirror parity |
| Fleet simulation | spanda fleet run — in-process multi-robot with deploy/peer wiring |
| Swarm coordinator (experimental) | swarm { fleet; policy; } + spanda swarm coordinate — round-robin cursors in .spanda/swarm-state.json; --mesh-url relays peer/leader-follow steps via fleet mesh |
| Robotics platform | mission, fleet, safety_zone, certify; navigation/fusion runtime; Nav2 adapter hook |
| OTA deploy CLI | spanda deploy plan|rollout|rollback|status — local rollout state (.spanda/deploy-state.json) |
| Remote OTA agents | spanda deploy agent start|register|list + deploy rollout --remote — HTTP agent on devices; --require-certify on agent and rollout |
| Fleet orchestration | spanda fleet orchestrate — round-robin mission coordination report; --remote relays peer steps via HTTP fleet agents |
| Fleet peer agents | spanda fleet agent start|register|list — on-device peer relay server (.spanda/fleet-agents.json) |
| Fleet mesh coordinator | spanda fleet mesh start + fleet orchestrate --mesh-url — centralized multi-host peer relay |
| Adapter package verify | spanda verify-adapter — validate [adapter] provides/requires against registry metadata |
| Tooling | Native CLI (check, verify, run, sim, fleet, deploy, fmt, lint, doc), package manager (init, build, test, install), prebuilt installable packages (Linux/macOS/Windows via GitHub Releases) |
| Showcase demos | spanda demo {rover,safety,verify,fleet,health,readiness,assurance,self-healing,continuity,adas}; examples/showcase/*; examples/solutions/adas/ (Stable — stable-hardening-adas.md) |
| Security / audit | Capabilities, secrets, signed messages, audit records |
| Secure communication | secure_comm, encrypted buses, trusted-source publish/receive enforcement, AES-GCM wire frames, TLS session + rustls PEM validation, spanda security check|audit, TS runtime parity |
| Digital twins | twin, mirror fields, replay buffer, twin sync telemetry |
| Real-time contracts | deadline, jitter <=, priority, critical isolated on tasks; latency pipeline budgets |
| Reliability runtime | Watchdogs, operating mode blocks, recover from, retry/fallback; topic QoS deadline detection |
| Runtime fault detection | heartbeat, memory_watch, resource_watch, restart_policy, on runtime crash; CLI spanda fault scan|report, spanda runtime health|diagnose, spanda replay --show-faults; mission trace fault frames |
| Mission trace replay | spanda sim --record, spanda replay, --deterministic, --playback, --wall-clock |
| Persistent telemetry | --persist-telemetry, SPANDA_TELEMETRY_STORE=1, spanda telemetry — JSONL or SQLite; OTLP push/serve, fleet-push mesh aggregation, sessions + replay |
| First-class regex | Literals, Regex type, string methods, trigger/subscribe filters, validate rules |
| Lean-core workspace | 50+ focused Rust crates; spanda-core facade; CLI/bindings use workspace deps directly (crates/README.md) |
| Verification & DX | spanda-capability — traceability, minimum-hardware, health analysis; spanda-readiness — operational readiness, mission verification, safety reports; spanda check --verification-json; LSP verification diagnostics and quick-fixes |
| Health & kill switch | health_check, health_policy, fleet require runtime; kill_switch, remote_signed, on kill_switch handlers |
| Self-healing & recovery | recovery_policy, recovery planner, validation gates, audit/traceability; runtime dispatch (modes, speed caps, fleet mesh relay, reassign → continuity mesh); CLI heal, recover, recovery-report, recovery knowledge, sim --inject-failure; fleet agent interpreter + assurance recovery; mission operator approval gating |
| Mission continuity | Checkpoint resume, state transfer, succession ranking, takeover/delegation; CLI continuity, takeover, delegate, succession; continuity_policy; diagnostics in spanda check --readiness-json; spanda demo continuity; official spanda-mission-continuity package |
| Distributed decisions | Brain/spinal-cord/reflex layers; decision_tree, offline_policy, entity local_decision_authority |
| Runtime policy gates; signed offline policy + signed decision trees + cache; live v3 signed trace emission | |
Persistent escalation store; runtime conflict resolution; CLI spanda decision *; attack simulations |
|
| Rule enforcement tests — Stable (stable-hardening-distributed-decisions.md) | |
| Differentiation NOW | spanda contract verify, spanda explain, spanda audit decisions, spanda safety-coverage, spanda recovery-coverage; spanda demo differentiation; decision trail showcase (differentiation/decision_trail/) with explain decision on v3 trace |
| What-If Analysis | spanda what-if, /v1/analytics/what-if, gRPC GetAnalyticsWhatIf — Stable — stable-hardening-what-if.md |
| Mission Risk Analysis | spanda risk, /v1/analytics/mission-risk — Stable — stable-hardening-mission-risk.md |
| Readiness Forecasting | spanda readiness forecast, /v1/analytics/readiness-forecast — Stable — stable-hardening-readiness-forecast.md |
| Trust Framework | spanda trust, /v1/trust/program, gRPC GetTrustProgram — Stable — stable-hardening-trust-framework.md |
| Trust Graph | spanda trust-graph, /v1/analytics/trust-graph — Stable — stable-hardening-trust-graph.md |
| Scorecards | spanda score, /v1/executive/scorecard — Stable — stable-hardening-scorecards.md |
| Digital Mission Twin | spanda twin mission — Stable — stable-hardening-digital-mission-twin.md |
| Certification Packs | spanda certify pack --bundle — Stable — stable-hardening-certification-packs.md |
| Mission Time Travel | spanda replay --at / --inspect — Stable — stable-hardening-mission-time-travel.md |
| Human/Robot Teaming | spanda team verify — Stable — stable-hardening-human-robot-teaming.md |
| Autonomous Governance | spanda governance (program policy blocks) — Stable — stable-hardening-autonomous-governance.md |
| Operational Governance | Autonomy levels, deployment profiles, certification lifecycle, risk, accountability, standards profiles |
Live decision enforcement; spanda compliance check, governance validate|report, certification list|inspect|report |
|
deployment profile|verify, risk report; REST/gRPC /v1/governance/*; Control Center Governance tab — Stable |
|
| governance.md · stable-hardening-operational-governance.md | |
| Twin Cloud SaaS | spanda twin cloud, /v1/twins/*, gRPC twins (proto 1.0.10), file-backed store + history — Stable — twin-cloud.md · stable-hardening-twin-cloud-saas.md |
| LATER Control Center analytics | /v1/analytics/{mission-twin,certification-pack,time-travel,human-teaming,governance} + gRPC GetAnalytics* (proto 1.0.7); Analytics tab |
| Mission assurance | knowledge_model, state_estimator, anomaly_detector, on anomaly, prognostics, mitigation, resilience_policy, assurance_case; CLI assure, anomaly scan, diagnose, state estimate, prognostics, mission verify, resilience check, mitigation plan; spanda demo assurance |
| Weighted sensor fusion | observe { }, state_estimator, fusion.read() with type-weighted confidence; spanda-fusion package |
| Learned anomaly runtime | learned backend assurance.anomaly; EMA volatility; optional ONNX (SPANDA_ANOMALY_ONNX_MODEL_PATH) |
| Typed handler I/O | Return types on behavior, task, trigger, event, and agent plan handlers (Rust + TS mirror) |
| Area | Capabilities | Caveats |
|---|---|---|
| Digital twins (live sync) | Twin mirror + replay; Twin Cloud SaaS (spanda twin cloud, /v1/twins/*, file-backed store + history, RBAC mutations) — Stable; legacy SPANDA_CLOUD_UPLOAD_URL + import-replay bridge |
Hosted managed product (billing, SLA) — hosted-twin-cloud-product.md |
| Replay | replay true, frame buffer, mission traces |
In-process only; v2 traces embed state snapshots for --playback |
| Advanced verification | Fault injection, compatibility matrix | Matrix may report stub targets |
| Multi-agent systems | Agent-to-agent comm, fleet peer messaging | In-process mesh + HTTP fleet agent relay (fleet orchestrate --remote / --mesh-url) |
| OTA rollout | Deploy plan/rollout/rollback/status | Local state file + HTTP deploy agents; --require-certify blocks uncertified rollouts |
| Certification metadata | certify ISO13849 { level PLd; } |
Verify-only metadata; --strict-certify / --enforce-certify; spanda certify prove; deploy plan proof summary |
| Nav2 / SLAM packages | Registry adapter stubs + example packages | External Nav2/Gazebo/OpenCV not bundled; optional SPANDA_NAV2_CMD / SPANDA_SLAM_CMD bridges |
| ROS2 adapter | Native rclrs cdylib, rclpy daemon, CLI bridge |
Requires ROS Humble; not default transport |
| LLVM / native codegen | spanda ir, llvm-ir, compile-native; scripts/llvm_golden_path.sh |
Early stage; not primary execution path |
| FFI | extern python/extern cpp subprocess bridges; optional cpp-native in-process |
PyO3 path is Tier 2 adoption unlock |
| World models | world_model { } block parser; fusion.read() → belief hook; world_model.update / belief / export; Rust + TS typecheck parity |
Minimal belief buffer; see world_model_patrol.sd |
| Ledger / provenance | spanda-ledger provider → MockLedgerBackend |
Mock chain only; no production blockchain adapters |
| MQTT / DDS live | SPANDA_LIVE_MQTT=1, --features live-mqtt; CI Nightly mqtt-golden-path |
DDS is UDP JSON shim, not full DDS middleware |
| Self-hosting bootstrap | examples/self_host/lexer_keywords.sd; Rust parity tests |
Rust compiler remains authoritative |
| LSP | Diagnostics, completion, hover, rename, verification quick-fixes | Requires built native CLI; VS Code extension with bundled LSP; continuity/recovery policy quick-fixes; CI builds VSIX |
| DAP debugger | Breakpoints, step over/in/out, every trigger entry |
VS Code + spanda-dap; tested in phase34_gaps.rs / phase35_gaps.rs |
| WASM / web playground | Browser check/run/verify | Limited surface vs native CLI |
| Live AI providers | OpenAI, Anthropic, ONNX via Python bridge | Requires API keys or SPANDA_ONNX_MODEL_PATH; mock fallback by default |
| Live IoT bridges | Modbus TCP, OPC-UA, zigbee, lora, matter, canbus | Env-gated (SPANDA_LIVE_*=1); in-memory hub fallback |
| Package publish | spanda publish, registry search, mirror to registry/packages/ |
Remote upload via SPANDA_REGISTRY_URL; hosted index lists 40 packages after build-registry.sh |
| Official package provenance | Registry-only provider bootstrap; path/git name-squatting blocked | OfficialProvenance API; official_provenance install warning; production deploy gate hard-fail |
| Registry signature policy | SPANDA_REGISTRY_REQUIRE_SIGNATURE=1 + lockfile signature audit |
Required for production deploy gate; optional at install otherwise |
| Area | Description |
|---|---|
| Human Interaction & Spatial Computing (H1–H6) | Stable — H1–H6 platform APIs, Control Center Humans tab, /v1/humans + /v1/hri/*, promotion gate hri_stable_promotion_gate.sh; registry vendor packages remain Experimental — stable-hardening-human-interaction.md |
| Smart Spaces & Ambient Intelligence | Stable — blueprint #15 feature-complete: simulation matrix, Control Center panels (REST + gRPC 1.0.5), live env I/O bridges, weighted readiness, CI smart_spaces_promotion_gate.sh — stable-hardening-smart-spaces.md · solutions/smart-spaces.md |
| Platform maturity (Phase A) | Stable — spanda graph, spanda deploy gate, spanda explain, spanda trust — platform-maturity-roadmap.md |
| Platform maturity (Phase B) | Stable — threat model, mission diff, scorecard (spanda score), policy engine (spanda verify --policy, readiness --policy, deploy gate --operational-policy, runtime --enforce-policy) |
| Platform maturity (Phase C) | Stable — chaos, readiness trends, resource estimation, compliance profiles, ADR (spanda adr) |
| Platform maturity (Phase D) | Stable — verify-time tamper/integrity, composite program trust, secure-boot attestation, compliance accreditation export, decision explain, runtime policy, AI generate/suggest, spoof-check, security assurance, tamper_policy runtime |
| Enterprise operations (E1–E4) | Control Center (spanda control-center serve, embedded UI, ControlCenterPanel in @davalgi-spanda/web, Tauri desktop 0.4.2) |
REST v1 (spanda-api); Device Pool lifecycle; host-backed discovery + pool ingest; RBAC v1 (SPANDA_API_KEY) |
|
ManagedSecretVault, alerting core (spanda-ops); provisioning/snapshots/discovery (E2) |
|
| Operational drift/OTA/trust/SRE/operator APIs + Python SDK + WebSocket telemetry + OTLP trace export (E3) | |
| Compliance export/digital thread/executive scorecard/PDF reports (E4); Stable tier | |
| enterprise-operations-roadmap.md · control-center.md | |
| Enterprise operations (organizational gates) | 30-day field soak completion (enterprise_ops_field_soak_init.sh); third-party security audit sign-off — tracked separately from tier promotion — enterprise-ops-stable-promotion.md |
| LLVM backend (production primary) | Optimized native binaries replacing interpreter as default deploy path |
| Self-hosting compiler (full) | Complete Spanda-authored compiler pipeline |
| ROS2 production adapter | First-class, zero-config ROS2 deployment |
| VS Code Marketplace publish | VSIX CI + local install ready; marketplace listing pending maintainer VSCE_PAT |
| Production blockchain | spanda-ledger-ethereum and related chain adapters |
| Full world models | Knowledge graphs, beliefs, policies beyond minimal runtime |
| Twin cloud SaaS (hosted product) | Managed multi-tenant service (billing, SLA, multi-region) — OSS /v1/twins/* is Stable; pilot at deploy/twin-cloud-hosted/ — hosted-twin-cloud-product.md |
See tier-3-experimental.md and tier-3-golden-paths.md.
| Feature | Replacement | Notes |
|---|---|---|
| Legacy inference-only AI paths | ai_model + agent |
Import-based ONNX/TFLite remain for classical workflows |
| TypeScript-only verification | Native spanda verify |
TS mirror validates deploy syntax; Rust CLI is authoritative |
spanda_core::transport_live |
spanda_transport_routing::transport_live |
Removed Phase 17 |
spanda_core::transport_mqtt / transport_dds / transport_websocket / transport_live |
spanda-transport-* or spanda_transport_routing::live_bridges |
Removed Phase 17 |
spanda_core::transport / transport_wire / transport_security / transport_rclrs |
spanda-transport-routing, spanda-transport, spanda-transport-ros2 |
Removed Phase 19 |
| Feature | Status | Notes |
|---|---|---|
| Lexer / parser / AST | Stable | Rust authoritative; TS mirror |
| Type checker + units | Stable | Physical unit algebra enforced |
| modules / imports | Stable | spanda install vendor support |
| structs / enums / traits | Stable | Generic struct literals; enum payloads |
| generics | Experimental | Module fn + struct type params |
trait objects (dyn Trait) |
Stable | Rust + TS mirror |
| match / Result / Option | Stable | |
| async / await | Stable | Cooperative single-threaded |
| spawn / select / channels | Stable | Cooperative concurrency with TS mirror |
triggers (on / every / when) |
Stable | Unified TriggerRegistry; see docs/triggers.md |
| test blocks | Stable | Rust runtime + TS runTests() |
extern fn / FFI |
Experimental | Subprocess bridges; optional in-process |
| Spanda IR (SIR) | Stable | JSON export via spanda ir |
| Codegen / LLVM | Experimental | HAL profiles; conditional codegen |
| Feature | Status | Notes |
|---|---|---|
| robot / sensor / actuator | Stable | |
| agent / goal / task / skill | Mock-backed (Stable API) | Default mock AI backend; live providers optional |
| ActionProposal → SafeAction | Stable | Compile + runtime |
| safety zones / emergency stop | Stable | |
| deterministic scheduler | Stable | task every Nms |
| deadline / jitter / priority | Stable | Compile-time validation; runtime telemetry |
| pipelines / watchdogs / modes | Stable | See docs/reliability.md, docs/watchdogs.md, docs/degraded-modes.md |
| mission trace replay | Stable | --record, spanda replay --deterministic / --playback |
| persistent telemetry store | Stable | --persist-telemetry, spanda telemetry; JSONL (default) or SQLite; OTLP export/push/serve, fleet-push, sessions — telemetry-store.md |
| regex literals / filters | Stable | See docs/regex.md |
| state machine / events | Stable | |
| twin / replay | Experimental | Replay buffer; live sync simulated |
| observe / fusion | Stable | Weighted fusion by sensor type; state_estimator runtime bindings |
| mission assurance (static + CLI) | Stable | spanda-assurance crate; 9 official packages (includes spanda-mission-continuity) |
| self-healing & recovery (static + CLI) | Stable | Recovery planner, validation gates, audit, knowledge store |
| Recovery Orchestrator | Stable | spanda-recovery crate; escalation levels 0–8, graph, playbooks, policies, simulation |
Predictive indicators, persisted history; CLI spanda recovery *; REST /v1/recovery/* (14 routes) |
||
gRPC (proto 1.0.14 via GET /v1/version); plugin [recovery.extensions]; SDK; Control Center Recovery tab |
||
scripts/recovery_orchestrator_smoke.sh, stable promotion gate — recovery-orchestrator.md |
||
| mission continuity (static + CLI + diagnostics) | Stable | spanda-assurance continuity module; CLI continuity, takeover, delegate, succession; continuity:* diagnostics in check JSON and LSP |
| mission continuity runtime dispatch | Stable | Interpreter mode-specific takeover, durable checkpoints, auto-trigger on health faults, fleet agent /v1/continuity/execute, mesh relay, swarm --failed handoff |
| self-healing runtime dispatch | Stable | Auto-trigger on health faults, approval polling/retry, fleet mesh relay with failure events, mission approval gating; scripts/fleet_field_validation.sh |
| fleet agent interpreter recovery | Stable | POST /v1/recovery/execute with recovery_engine: interpreter; scripts/fleet_agent_recovery_smoke.sh |
| recovery diagnostics (CLI + LSP) | Stable | spanda check --readiness-json merges recovery:* categories; TS mirror in scripts/lsp-readiness.mts |
| continuity diagnostics (CLI + LSP) | Stable | spanda check --readiness-json merges continuity:* categories including continuity:mission; TS mirror in src/continuity-diagnostics.ts |
| learned anomaly backends | Experimental | Runtime scan_learned; ONNX optional |
| verify { } behavioral assertions | Stable | |
| hardware / deploy | Stable | Rust verify CLI |
| Feature | Status | Notes |
|---|---|---|
| Native CLI (full) | Stable | check, verify, run, sim, replay, fleet, fmt, lint, doc, man, reference, package |
| Prebuilt packages | Stable | Linux/macOS/Windows archives, shell/PowerShell installers, Windows MSI, Homebrew formula; see installation.md |
| TypeScript CLI | Stable | Delegates to Rust when built |
| Formatter / linter / docgen | Stable | /// doc comments in .sd; spanda doc (markdown/HTML/JSON); spanda man; man pages |
| LSP | Experimental | VS Code extension scaffold; CI builds VSIX on push |
| DAP debugger | Experimental | VS Code + spanda-dap; every trigger entry (Phase 35) |
| N-API | Experimental | check, run, verify, sir, fmt |
| WASM | Experimental | check, run, verify, sir, fmt |
| Feature | Status | Notes |
|---|---|---|
| python.* / cpp.* imports | Experimental | Subprocess bridges |
| ROS2 adapter | Experimental | Native rclrs cdylib; CI Nightly ros2-rclrs-native on Ubuntu 22.04 + Humble |
| Transport adapters | Experimental | In-memory + optional rclrs/rclpy |
| Package manager | Stable | Hosted index + local mirror; spanda publish copies to registry/packages/ |
| LLVM / native codegen | Experimental | compile-native early stage |
| Pillar | Status | Key surfaces |
|---|---|---|
| Control Center | Stable | spanda control-center serve, embedded HTML + @davalgi-spanda/web panel (RBAC-aware tabs, sidebar version, Administration user/key/alert config, Mission/Simulation/Replay, drift, trust, compliance, SRE, executive); Tauri desktop 0.6.3 (desktop-v0.6.3) — control-center-versioning.md |
| Device Pool | Stable | Lifecycle states, assign/trust/quarantine/retire, failover chains; multi-tenant API key scoping |
| Device Discovery | Stable | Subnet, mDNS, BLE, USB, CAN, MQTT, ROS2 host probes; production TLS policy (SPANDA_DISCOVERY_REQUIRE_TLS, spanda-discovery-tls) |
| Provisioning | Stable | POST /v1/provision, discover → ready workflow |
| Configuration Management | Stable | Snapshots, diff, resolve; approval queue + publish-on-approve (/v1/config/approvals) |
| RBAC | Stable | 7 roles, SPANDA_API_KEY, /v1/rbac/matrix |
| Secret Management | Stable | ManagedSecretVault, rotation metadata |
| Telemetry | Stable | Health/readiness/mission signals; trend analysis; forecasting Planned |
| Alerting | Stable | Webhook, email, PagerDuty (bi-directional sync), Teams; per-severity dedup; registry alert packages |
| Configuration Drift | Stable | Full operational drift API; scheduled scans (SPANDA_DRIFT_SCAN_INTERVAL_SECS); seven dimensions via GET /v1/drift |
| OTA & Rollback | Stable | Canary, blue/green, phased dry-run; production --require-certify via SPANDA_OTA_REQUIRE_CERTIFY |
| Package Trust | Stable | spanda trust, /v1/trust/package, trust score |
| SDKs | Stable | Official Rust/Python/TypeScript clients published at 0.4.2 (cargo add spanda-sdk, pip install spanda-sdk, npm install @davalgi-spanda/sdk) — entity read/eval/mutation helpers (entityReadiness, entityRelationships, gRPC entity_health/entity_trust); @davalgi-spanda/web Control Center panel; program-level REST + gRPC; legacy packages/sdk-python |
| Operator Workflows | Stable | Mission approve, takeover, quarantine, recovery approval |
| SRE | Stable | /v1/sre/summary with slo, burn_rate, MTTR/MTBF hints; incident workflow; PagerDuty webhook sync; fast-burn background monitor |
| Reporting | Stable | Markdown, JSON, PDF exports; scheduled webhook delivery (GET/POST /v1/reports/schedules) |
| Compliance | Stable | Evidence packs, GET /v1/compliance/export, signed profile catalog (GET /v1/compliance/profiles) |
| APIs | Stable | REST v1 + OpenAPI; unified entity routes (/v1/entities/*); program-level SDK routes (/v1/programs/*) |
JSON-RPC gateway (Control Center methods); native gRPC (tonic) — 164 RPCs, proto 1.0.14 via GET /v1/version |
||
Rate limits (SPANDA_API_RATE_LIMIT_PER_MINUTE); versioning (GET /v1/version, X-Spanda-Api-Version) |
||
| Unified Entity Model | Stable | EntityRegistry projects fleet, devices, humans, providers, packages into entity graph |
Verification, readiness, health, and trust via verify_entity / evaluate_entity_* |
||
Control Center Entities tab with read/write mutations; CI entity_model_smoke.sh (REST + TS + Python + Rust SDK) |
||
| SDKs 0.4.2 on crates.io, PyPI, npm — entity-model.md, entity-apis.md, entity-sdk.md | ||
| Observability | Stable | OTLP trace/metrics export, correlation IDs, WebSocket telemetry; spanda-otel-collector; Grafana templates (spanda-grafana-dashboards); HA collector guide |
| Digital Thread | Stable | Full lifecycle graph (requirement → retirement); GET /v1/digital-thread/query with lifecycle_phase filter |
See enterprise-operations-roadmap.md · control-center.md · stable-hardening-enterprise-ops.md
| Area | Status | Key surfaces |
|---|---|---|
| ADAS & Autonomous Driving | Stable | spanda demo adas, ISO 26262 profile, golden traces, Control Center ADAS tab — stable-hardening-adas.md |
| Human Interaction & Spatial Computing | Stable | H1–H6 APIs, Humans tab, /v1/humans, /v1/hri/*, spatial_computing_smoke.sh — stable-hardening-human-interaction.md |
| Smart Spaces & Ambient Intelligence | Stable | Six blueprint apps, Smart Spaces tab, BACnet/KNX/HA bridges, smart_spaces_promotion_gate.sh — stable-hardening-smart-spaces.md |
| Platform maturity (Phases A–D) | Stable | Graph, deploy gate, explain, trust, threat model, policy engine, chaos, scorecard, tamper, compliance export, ADR — platform-maturity-roadmap.md |
Stable promotion (2026-07-02): Human Interaction & Spatial Computing (H1–H6), Smart Spaces blueprint, ADAS blueprint, and Platform maturity Phases A–D promoted to Stable after implementation promotion gates (soak/audit skipped in CI). Organizational gates — per-blueprint field soak and third-party security audit sign-off — remain tracked in stable-hardening-human-interaction.md, stable-hardening-smart-spaces.md, stable-hardening-adas.md, and enterprise-ops-stable-promotion.md.
Enterprise ops (2026-06-28): All 20 enterprise operations pillars promoted to Stable after
enterprise_ops_stable_promotion_gate.sh (implementation checks). SDK 0.4.2 and desktop
0.4.2 (desktop-v0.4.2) published. Ongoing organizational gates — 30-day field soak
(field-soak-gate.md) and third-party security audit sign-off — tracked in
enterprise-ops-stable-promotion.md.
Release-hardening pass against the strict labels above (#50):
| Area | Label | Rationale |
|---|---|---|
| AI agents / LLM providers | Mock-backed (Stable API) | Default path is mock; live keys optional |
| Live IoT / MQTT / ROS2 / LLVM | Experimental | Env-gated or optional toolchains |
| Ledger / blockchain | Stubbed / Mock-backed | MockLedgerBackend only |
| Solution blueprints (ADAS, Smart Spaces, HRI) | Stable implementation; organizational soak open | CI promotion gates skip field soak / third-party audit (#51) |
| Enterprise ops pillars | Stable implementation; organizational soak open | Same |
Remaining review: any row still marked Stable whose default path is simulated-only or docs-only should be demoted in a follow-up PR.
OPENAI_API_KEY, ANTHROPIC_API_KEY, or
SPANDA_ONNX_MODEL_PATH for live calls (SPANDA_LIVE_AI=0 forces mock). Label: Mock-backed.spanda publish mirrors bundles to registry/packages/ in-repo; remote upload requires
SPANDA_REGISTRY_URL. Run ./scripts/build-registry.sh to refresh the hosted index after adding
scaffolds under packages/registry/.vscode-extension-ci.yml;
Marketplace publish pending maintainer VSCE_PAT.spanda fleet mesh start, --mesh-url on
orchestrate/swarm).| Item | Category | Detail |
|---|---|---|
| Global package registry | Hosted + mirror | Default SPANDA_REGISTRY_URL points at repo index; spanda publish mirrors to registry/packages/ |
| Live OpenAI / Anthropic / ONNX | Optional live path | OPENAI_API_KEY, ANTHROPIC_API_KEY, or SPANDA_ONNX_MODEL_PATH; Python bridge; mock fallback |
| Live Modbus / OPC-UA IoT | Optional live hardware | SPANDA_LIVE_MODBUS=1 / SPANDA_LIVE_OPCUA=1; --features live-iot for native Modbus TCP |
| IoT protocol bridges (zigbee/lora/matter/canbus/bacnet/knx) | Live + hub fallback | SPANDA_LIVE_ZIGBEE=1 etc.; BACnet/KNX via bacpypes3/xknx or SPANDA_*_CMD; ./scripts/live_iot_golden_path.sh, ./scripts/smart_spaces_live_iot_smoke.sh |
| Kill switch remote_signed | Runtime + verify enforced | Requires kill_switch_signature JSON when remote_signed is set; verify reports error without signed comm |
| MQTT / DDS / WebSocket live transport | Production wire + optional live brokers | AES-256-GCM wire frames; live MQTT/WebSocket/DDS via --features live-transport + SPANDA_LIVE_MQTT=1 / SPANDA_LIVE_WEBSOCKET=1 / SPANDA_LIVE_DDS=1; TypeScript mirrors the same env flags |
| Secure comm live crypto | Production wire | AES-256-GCM for transport frames and EncryptedMessage payloads; session material from robot secrets |
| Full native binary deploy | Experimental | spanda deploy --target native, compile-native (clang + llvm feature) |
| Blockchain / ledger cloud | Stubbed | Audit records local; see future-blockchain-support.md |
Nothing in the Supported list above is known broken in CI Fast (./scripts/ci-fast.sh) or CI
Integration on main. ROS2 rclrs native and cargo audit run in CI Nightly. See
ci-architecture.md.
.sd source
→ lexer → parser → AST
→ type checker (+ units, safety, capabilities)
→ [optional] hardware verifier (deploy targets)
→ interpreter + simulator
→ [optional] SIR → LLVM (experimental)
| Crate | Role |
|---|---|
spanda-core |
Language implementation (authoritative) |
spanda-cli |
Native spanda binary |
spanda-package |
Package manager |
spanda-audit / spanda-security |
Audit and security |
spanda-llvm / spanda-rt |
Experimental native codegen |
spanda-node / spanda-wasm |
Bindings |
spanda-dap |
Debug adapter |
@spanda/lsp / @davalgi-spanda/web |
LSP and web playground |
See architecture.md for diagrams.
--verification-json and LSP
quick-fixesexpect_compile_error and test CLI