diff --git a/CHANGELOG.md b/CHANGELOG.md index da56654..ccd9e2a 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -7,6 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0 ## [Unreleased] +### Changed + +- **Documentation sweep ([#14](https://github.com/ARPAHLS/aura/issues/14))** — `docs/INDEX.md` three-tier entry (Start / Build / Decide + Optional vision); demoted narrative, three-rings, aura-levels, field-services; refreshed architecture, concepts, stack-position, field-services shipped vs planned; fixed stale v0.2 voice in getting-started and concepts. + ## [0.3.4] - 2026-08-26 ### Added diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 2bb7a4b..e318dc4 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -159,7 +159,8 @@ Pure internal refactors with no user-visible effect may omit CHANGELOG; ask on t | Integration example (Ollama, API, framework) | `integrations//`, `docs/integrations/README.md`, `.env.example`, CHANGELOG | | CLI commands or flags | `docs/getting-started.md`, `README.md` quick start line, `docs/outputs.md` (export/compare shapes), CHANGELOG | | New core example | Flat script in `examples/`, `examples/README.md`, optional link from `docs/getting-started.md` | -| Architecture terminology | `docs/architecture.md`, `README.md` diagrams (keep in sync) | +| Architecture terminology | `docs/architecture.md`, `docs/INDEX.md`, `README.md` diagrams (keep in sync) | +| Public docs tree / INDEX tiers | `docs/INDEX.md`, cross-links from `getting-started.md`, `onboarding.md`, `ROADMAP.md` | | Release / PyPI behavior | `docs/PUBLISHING.md`, `.github/workflows/publish-pypi.yml`, CHANGELOG | | PR CI workflow | `.github/workflows/ci.yml`, `.github/workflows/reusable-test.yml`, `docs/TESTING.md`, `CONTRIBUTING.md` | | Issue template fields | `.github/labels.json` if new label needed; run label sync | diff --git a/README.md b/README.md index 94450ae..4ca0950 100644 --- a/README.md +++ b/README.md @@ -122,6 +122,7 @@ CLI: `aura agent create`, `aura run`, `aura export`, `aura report show`, `aura c | Topic | Links | | :--- | :--- | +| **Index** | [docs/INDEX.md](docs/INDEX.md) — full doc map (Start / Build / Decide) | | **Start** | [onboarding.md](docs/onboarding.md) · [getting-started.md](docs/getting-started.md) · [concepts.md](docs/concepts.md) · [using-aura.md](docs/using-aura.md) | | **Integration** | [reference-tool-host-capstone.md](docs/guides/reference-tool-host-capstone.md) · [guides/aura-on-skillware.md](docs/guides/aura-on-skillware.md) · [skillware-integration.md](docs/skillware-integration.md) · [sequencer.md](docs/sequencer.md) | | **Identity & audit** | [trust-paths.md](docs/trust-paths.md) · [outputs.md](docs/outputs.md) | diff --git a/docs/INDEX.md b/docs/INDEX.md index c19706c..90404b3 100644 --- a/docs/INDEX.md +++ b/docs/INDEX.md @@ -1,57 +1,69 @@ # AURA Harness — Documentation +Single entry point for public docs. **Shipped behavior** lives in Tier 1–2; Tier 3 is positioning; **Optional** is vision and internal vocabulary — not required for onboarding. + --- -## Start here +## Tier 1 — Start | Doc | Content | |---|---| | [onboarding.md](onboarding.md) | **Start here** — install → posture → agent → body → receipt | -| [getting-started.md](getting-started.md) | Install, example, CLI | -| [using-aura.md](using-aura.md) | Membrane, personas, SDK | -| [contributing/ai_native_workflow.md](contributing/ai_native_workflow.md) | Agents and operators | -| [TESTING.md](TESTING.md) | pytest, black, flake8, PR checklist | -| [../CONTRIBUTING.md](../CONTRIBUTING.md) | Full contributor guide (humans and agents) | +| [getting-started.md](getting-started.md) | Install, minimal example, CLI | +| [using-aura.md](using-aura.md) | Membrane, postures, session export, SDK | | [concepts.md](concepts.md) | Agent, session, identity, audit | -| [comparison.md](comparison.md) | vs orchestrators, eval harnesses, tracing | -| [architecture.md](architecture.md) | v0.2 modules | -| [ROADMAP.md](ROADMAP.md) | Shipped vs deferred | -| [examples/](../examples/README.md) | Four runnable core demos plus Skillware / ToolHost demos | -| [guides/reference-tool-host-capstone.md](guides/reference-tool-host-capstone.md) | ToolHost integration checklist (AURA-first) | +| [examples/README.md](../examples/README.md) | Runnable core scripts + integration demos | +| [integrations/README.md](integrations/README.md) | Pick your stack — Ollama, cloud APIs, Skillware | +| [guides/reference-tool-host-capstone.md](guides/reference-tool-host-capstone.md) | ToolHost checklist (mock → live → Ollama) | --- -## Integration +## Tier 2 — Build | Doc | Content | |---|---| -| [integrations/README.md](integrations/README.md) | Pick your stack — models, tools, frameworks | -| [skillware-integration.md](skillware-integration.md) | Skillware reference adapter — see `integrations/skillware/` | -| [sequencer.md](sequencer.md) | Prescriptive pipelines, gates, conformance | +| [architecture.md](architecture.md) | Modules, data flow, extension surface | +| [stack-position.md](stack-position.md) | Optional — harness-centric input layers vs full ARPA stack | +| [sequencer.md](sequencer.md) | Prescriptive pipelines, gates, `when`, conformance | +| [skillware-integration.md](skillware-integration.md) | Skillware reference adapter (optional host) | +| [guides/aura-on-skillware.md](guides/aura-on-skillware.md) | Deep dive — Skillware as one ToolHost impl | +| [trust-paths.md](trust-paths.md) | `agent_ref`, ULID, ids trailer — no central ID service | +| [outputs.md](outputs.md) | JSONL, summary, audit report, hash chain, OTel | +| [TESTING.md](TESTING.md) | pytest, black, flake8, PR checklist | +| [../CONTRIBUTING.md](../CONTRIBUTING.md) | Contributor guide (humans and agents) | +| [contributing/ai_native_workflow.md](contributing/ai_native_workflow.md) | Agent contribution workflow | + +Copy [`.env.example`](../.env.example) to `.env` for local Ollama or cloud API keys. Never commit `.env`. + +--- + +## Tier 3 — Decide -Copy [`.env.example`](../.env.example) to `.env` for local Ollama (`llama3.2:1b`) or cloud API keys. Never commit `.env`. +| Doc | Content | +|---|---| +| [comparison.md](comparison.md) | vs orchestrators, eval harnesses, tracing | +| [ROADMAP.md](ROADMAP.md) | Shipped vs deferred (sole deferrals file for public docs) | +| [README.md](../README.md) | Project entry, quick start, badges | --- -## Reference +## Optional — vision & reference + +Not on the default onboarding path. Kept for ARPA stack context and long-form design language. | Doc | Content | |---|---| -| [README.md](../README.md) | Project entry, stack diagram | -| [narrative.md](narrative.md) | Long-form vision (coat, SCI) | -| [stack-position.md](stack-position.md) | Optional ARPA stack context | -| [trust-paths.md](trust-paths.md) | Lite ID — no identity service | -| [outputs.md](outputs.md) | AuraEvent, exporters | -| [aura-levels.md](aura-levels.md) | Autonomy tiers (roadmap enforcement) | -| [field-services.md](field-services.md) | Observer presets (roadmap) | +| [narrative.md](narrative.md) | Long-form vision (coat, SoulSig, cybernetics) | | [three-rings.md](three-rings.md) | Envelope · Field · Adapter model | -| [glossary.md](glossary.md) | Terms | +| [aura-levels.md](aura-levels.md) | Autonomy tiers (enforcement roadmap) | +| [field-services.md](field-services.md) | Twelve parallel services — shipped vs planned | +| [glossary.md](glossary.md) | Terminology reference | --- ## Specifications -Schemas for manifest, events, plugins — **stable contracts** for adapters. +Stable contracts for adapters and tooling. | File | Purpose | |---|---| diff --git a/docs/ROADMAP.md b/docs/ROADMAP.md index 6f7e436..0ec8607 100644 --- a/docs/ROADMAP.md +++ b/docs/ROADMAP.md @@ -9,8 +9,9 @@ Shipped work stays in [CHANGELOG.md](../CHANGELOG.md). This file lists what is * | Version | Highlights | |---|---| | **v0.1** | Registry, sessions, constraints, JSONL export, SDK | -| **v0.2** | Membrane, sequencer, Skillware host, observer presets (monitor, break) | +| **v0.2** | Membrane, sequencer, Skillware host, observers | | **v0.3** | ULID + `agent_ref`, audit report, hash chain, OTel export (+ promoted attrs), compare CLI, ToolHost reference coat | +| **v0.3.4** | Onboarding guide, `report show`, flat core examples, Monitor/Break presets, capstone + examples 05–08 | --- diff --git a/docs/architecture.md b/docs/architecture.md index 97d987f..a7b248c 100644 --- a/docs/architecture.md +++ b/docs/architecture.md @@ -38,8 +38,12 @@ flowchart LR | `aura/observers/` | Parallel subscribers | | `aura/exporters/` | JSONL summary, OTel JSONL | -## Extension surface (roadmap) +## Extension surface -Type plugins, packaged observer presets (monitor, break — see [reference-tool-host-capstone.md](guides/reference-tool-host-capstone.md)), HTTP fleet API — see [ROADMAP.md](ROADMAP.md). +**Shipped:** ToolHost protocol, Monitor + Break observer presets, OTel export with promoted span attributes, sequencer `when` guards — see [reference-tool-host-capstone.md](guides/reference-tool-host-capstone.md). + +**Roadmap:** Limit preset, signed audit packs, HTTP fleet API, brain/memory/MCP type plugins — [ROADMAP.md](ROADMAP.md). + +Input-layer model (brain, memory, tools as adapters): [stack-position.md](stack-position.md). **Principle:** new capabilities emit or subscribe to the spine — core loop unchanged. diff --git a/docs/aura-levels.md b/docs/aura-levels.md index 943f4de..f7dda48 100644 --- a/docs/aura-levels.md +++ b/docs/aura-levels.md @@ -1,5 +1,7 @@ # AURA Levels +> **Optional / roadmap** — autonomy tiers; enforcement UX tracked in [#27](https://github.com/ARPAHLS/aura/issues/27). Today use postures in [using-aura.md](using-aura.md) and sequencer gates. + **Permissioned autonomy** — not binary on/off. From [narrative.md](narrative.md). Enforced by Spectrum + conformance engine + hook pipeline. diff --git a/docs/comparison.md b/docs/comparison.md index ea116ba..2f69e94 100644 --- a/docs/comparison.md +++ b/docs/comparison.md @@ -155,7 +155,7 @@ OpenTelemetry and structured logging provide **telemetry primitives** — spans, ### Key differences * **Agent semantics**: AURA events carry **agent identity**, session mode, constitution violations, and conformance — not generic spans alone. -* **Policy**: Logging does not enforce confirm-before-action or token budgets; AURA constraints do (v0.1 on explicit events; membrane intercept roadmap). +* **Policy**: Logging does not enforce confirm-before-action or token budgets; AURA constraints do on wired egress paths (broader intercept planned). * **Complementary**: OTel exporter shipped (v0.3); audit trail maps to spans ([outputs.md](outputs.md), [ROADMAP](ROADMAP.md)). --- diff --git a/docs/concepts.md b/docs/concepts.md index dd7d68a..4517640 100644 --- a/docs/concepts.md +++ b/docs/concepts.md @@ -4,11 +4,18 @@ Plain-language model for AURA Harness. ## Agent -A logical entity you run under AURA. Gets a permanent **`AURA-000n`** ID for audit. You can also give it a **name** (`agent1test`). +A logical entity you run under AURA. -Your own IDs (OpenAI assistant id, company id, etc.) live in the agent **`ids`** trailer — AURA does not replace them. +| Field | Role | +|---|---| +| **`agent_ref`** | Stable slug for humans and CI, e.g. `acme/compliance-bot` | +| **`aura_id`** | Internal ULID (default) or your supplied id | +| **`name`** | Optional alias for lookup | +| **`ids`** | Trailer for tenant, Skillware, and your external ids | + +Legacy profiles with `AURA-000n` ids still load. See [trust-paths.md](trust-paths.md). -v0.2 fields: **`skills`**, **`sequencer`** spec, **`observers`** list. +Profile fields also include **`skills`**, **`sequencer`** spec, **`observers`**, and **`rules`**. ## Session diff --git a/docs/field-services.md b/docs/field-services.md index a0dfe8c..92eb932 100644 --- a/docs/field-services.md +++ b/docs/field-services.md @@ -1,25 +1,27 @@ # Field Services +> **Optional reference** — not required for onboarding. See [INDEX.md](INDEX.md). For shipped observer behavior use [using-aura.md](using-aura.md) and [reference-tool-host-capstone.md](guides/reference-tool-host-capstone.md). + The **twelve complementary services** that run **in parallel** with the agent loop — the coat, not bolt-ons. -From [narrative.md](narrative.md). Implemented as **operation plugins** in `aura/ops/`, registered in [spec/capability.registry.json](../spec/capability.registry.json). +Design language from [narrative.md](narrative.md). Operation ids live in [spec/capability.registry.json](../spec/capability.registry.json); packaged presets ship under `aura/observers/presets/`. --- -| Service | What it does | -|---|---| -| **Monitor** | Loop state, tool calls, outputs, drift — continuously | -| **Audit** | Record what, when, why, under which permissions — **always on (Job B)** | -| **Track** | Task progress, resource use, retries, lineage across steps | -| **Limit** | Budgets, rate caps, scope, AURA Level permissions | -| **Safeguard** | Enforce guardrails from manifest, soul contract, memory layer, skill constitution, or elsewhere | -| **Wake** | Restart stalled loops, re-queue work, resume | -| **Break** | Stop infinite retries, circular reasoning, runaway tools | -| **Conserve** | Reduce token waste — redundant calls, repeated failures | -| **Recover** | Catch errors, retry logic, fallback paths | -| **Remember** | Memory discipline — keep, compress, discard, persist | -| **Learn** | Capture mistakes and outcomes for next iteration | -| **Attach** | Modular extensions — skills, schedulers, observers | +| Service | What it does | Status | +|---|---|---| +| **Monitor** | Loop state, tool calls, outputs, drift — continuously | **Shipped** — `preset: monitor` | +| **Audit** | Record what, when, why, under which permissions — always on | **Shipped** — audit spine + export | +| **Break** | Stop infinite retries, circular reasoning, runaway tools | **Shipped** — `preset: break` | +| **Track** | Task progress, resource use, retries, lineage across steps | Planned | +| **Limit** | Budgets, rate caps, scope, spectrum permissions | Planned | +| **Safeguard** | Enforce guardrails from manifest and constitution | Partial — constraint engine + manifest merge | +| **Wake** | Restart stalled loops, re-queue work, resume | Planned | +| **Conserve** | Reduce token waste — redundant calls, repeated failures | Planned | +| **Recover** | Catch errors, retry logic, fallback paths | Partial — sequencer retries | +| **Remember** | Memory discipline — keep, compress, discard, persist | Planned (memory adapter) | +| **Learn** | Capture mistakes and outcomes for next iteration | Planned | +| **Attach** | Modular extensions — skills, schedulers, observers | **Shipped** — observers + Skillware host | --- @@ -35,14 +37,14 @@ All three emit to the same **audit spine**. --- -## Spectrum Toggle +## Spectrum toggle (roadmap) -Manifest `spectrum.services` selects which field services activate for a session. `audit` is non-optional in production profiles. +Manifest `spectrum.services` will select which field services activate for a session. `audit` is non-optional in production profiles. Enforcement UX: [#27](https://github.com/ARPAHLS/aura/issues/27). --- -## Attachments (Extensions) +## Attachments (extensions) Beyond core twelve — observability modules, resource governors, temporal schedulers, event bridges, recovery playbooks. Registered as **op plugins** or **type plugins** — same extensibility model. -See [architecture.md](architecture.md) · [outputs.md](outputs.md) · [spec/capability.registry.json](../spec/capability.registry.json) +See [architecture.md](architecture.md) · [outputs.md](outputs.md) · [ROADMAP.md](ROADMAP.md) diff --git a/docs/getting-started.md b/docs/getting-started.md index d8e9827..d8e9fd0 100644 --- a/docs/getting-started.md +++ b/docs/getting-started.md @@ -47,7 +47,7 @@ with ag.session() as run: print(run.exports) # JSONL + summary paths ``` -## Sequencer example (v0.2) +## Sequencer example ```python from aura import agent, configure @@ -74,7 +74,7 @@ print(run.exports) Logs land in `~/.aura/sessions/` unless you configure project storage. -## CLI +## CLI (export & verify) ```bash aura agent create my-bot --purpose "research assistant" @@ -127,5 +127,5 @@ See [examples/](../examples/README.md) — including [sequencer_pipeline.py](../ - [skillware-integration.md](skillware-integration.md) — Skillware host - [concepts.md](concepts.md) — agent, session, sequencer - [comparison.md](comparison.md) — vs orchestrators and eval harnesses -- [ROADMAP.md](ROADMAP.md) — what comes after v0.2 +- [ROADMAP.md](ROADMAP.md) — shipped vs deferred - [architecture.md](architecture.md) — modules and data flow diff --git a/docs/glossary.md b/docs/glossary.md index 8059112..5ebea40 100644 --- a/docs/glossary.md +++ b/docs/glossary.md @@ -1,5 +1,7 @@ # Glossary +> **Optional reference** — terminology including ARPA-internal names. Product path: [concepts.md](concepts.md) and [INDEX.md](INDEX.md). + | Term | Meaning | |---|---| | **αύρα / AVRA / AURA** | Runtime coat / harness / field around the loop | diff --git a/docs/guides/README.md b/docs/guides/README.md index 0202ef3..b24b2e4 100644 --- a/docs/guides/README.md +++ b/docs/guides/README.md @@ -6,6 +6,6 @@ Long-form integration and usage guides. |---|---| | [aura-on-skillware.md](aura-on-skillware.md) | Reference adapter deep dive (Skillware as one ToolHost impl) | | [reference-tool-host-capstone.md](reference-tool-host-capstone.md) | **360° tool-host checklist** — membrane before/at/after, spine events, runnable paths | -| [skillware-follow-ups.md](skillware-follow-ups.md) | Post–#12 backlog (CI, flat examples, limit preset) | +| [skillware-follow-ups.md](skillware-follow-ups.md) | Post–#12 backlog (CI, limit preset, docs sweep) | See also: [using-aura.md](../using-aura.md), [skillware-integration.md](../skillware-integration.md). diff --git a/docs/guides/aura-on-skillware.md b/docs/guides/aura-on-skillware.md index 019002c..e6e3edf 100644 --- a/docs/guides/aura-on-skillware.md +++ b/docs/guides/aura-on-skillware.md @@ -273,6 +273,6 @@ See [skillware-follow-ups.md](skillware-follow-ups.md) for post–#12 backlog: f ## Related -- [using-aura.md](../using-aura.md) — membrane and personas +- [using-aura.md](../using-aura.md) — membrane and postures - [stack-position.md](../stack-position.md) — where AURA fits in the agent stack - [Skillware repository](https://github.com/arpahls/skillware) diff --git a/docs/guides/skillware-follow-ups.md b/docs/guides/skillware-follow-ups.md index 153df43..f4c21eb 100644 --- a/docs/guides/skillware-follow-ups.md +++ b/docs/guides/skillware-follow-ups.md @@ -25,7 +25,7 @@ Track these **after** closing the reference ToolHost epic ([#12](https://github. | **Flat examples restructure** ([#21](https://github.com/ARPAHLS/aura/issues/21)) | Align paths referenced in docs after examples layout change (remainder of [#41](https://github.com/ARPAHLS/aura/issues/41)) | | **Multi-provider comparison page** | One doc comparing Ollama vs GPT vs Claude vs Gemini with the same skill chain (extends capstone) | | **Skill catalog appendix** | Table of bundled Skillware skills: offline vs API, suggested AURA guardrails | -| **Docs sweep** ([#13](https://github.com/ARPAHLS/aura/issues/13), [#14](https://github.com/ARPAHLS/aura/issues/14), [#19](https://github.com/ARPAHLS/aura/issues/19), [#20](https://github.com/ARPAHLS/aura/issues/20)) | Cross-link INDEX, ROADMAP, integration READMEs | +| **Docs sweep** ([#14](https://github.com/ARPAHLS/aura/issues/14), [#19](https://github.com/ARPAHLS/aura/issues/19), [#20](https://github.com/ARPAHLS/aura/issues/20)) | Cross-link INDEX, ROADMAP, integration READMEs — onboarding ([#13](https://github.com/ARPAHLS/aura/issues/13)) shipped | --- diff --git a/docs/narrative.md b/docs/narrative.md index c40165e..801e608 100644 --- a/docs/narrative.md +++ b/docs/narrative.md @@ -1,5 +1,7 @@ # AURA Harness +> **Optional / vision** — long-form ARPA narrative. For product usage start at [onboarding.md](onboarding.md) and [INDEX.md](INDEX.md). + **The runtime coat around your agent loops.** *By ARPA Hellenic Logical Systems — [arpacorp.net](https://arpacorp.net)* diff --git a/docs/onboarding.md b/docs/onboarding.md index 60a13c9..c6d9ca7 100644 --- a/docs/onboarding.md +++ b/docs/onboarding.md @@ -1,6 +1,6 @@ # Onboarding — use AURA correctly -Step-by-step path from install to a reviewable session receipt. For membrane detail see [using-aura.md](using-aura.md); for terms see [concepts.md](concepts.md). +Step-by-step path from install to a reviewable session receipt. Full doc map: [INDEX.md](INDEX.md). For membrane detail see [using-aura.md](using-aura.md); for terms see [concepts.md](concepts.md). --- diff --git a/docs/outputs.md b/docs/outputs.md index 7c90c1b..33c3ea6 100644 --- a/docs/outputs.md +++ b/docs/outputs.md @@ -2,6 +2,8 @@ What a session produces on close (v0.3). +Identity fields on export: [trust-paths.md](trust-paths.md). Session-close workflow: [using-aura.md](using-aura.md). + --- ## Per session diff --git a/docs/skillware-integration.md b/docs/skillware-integration.md index 5556eea..05959eb 100644 --- a/docs/skillware-integration.md +++ b/docs/skillware-integration.md @@ -237,7 +237,7 @@ Use session export `.summary.json` `conformance.passed` in CI to fail builds whe ## Related -- [using-aura.md](using-aura.md) — membrane and personas +- [using-aura.md](using-aura.md) — membrane and postures - [sequencer.md](sequencer.md) — step model and gates - [integrations/skillware/](../integrations/skillware/) — reference scripts and README - [Skillware repo](https://github.com/arpahls/skillware) — skill registry, manifests, CLI diff --git a/docs/stack-position.md b/docs/stack-position.md index e8a7aae..3b4e2da 100644 --- a/docs/stack-position.md +++ b/docs/stack-position.md @@ -1,8 +1,8 @@ # AURA Architecture -How **AURA Harness** models a run — a harness-centric view of the broader ARPA stack. +> **Optional** — harness-centric view of the broader ARPA stack. Default path: [architecture.md](architecture.md) and [onboarding.md](onboarding.md). -The full stack diagram covers Identity → Soul → Body → Aura → Rooms / Legacy. **This repo uses a harness-centric view:** parallel inputs, one body, one coat, one audit trail out. +How **AURA Harness** models a run. --- @@ -25,17 +25,17 @@ flowchart LR ## Layers -| Layer | What it is | v0.1 | +| Layer | What it is | Shipped today | | :--- | :--- | :--- | -| **Identity** | `AURA-000n`, optional name, `ids.external` trailer | Shipped — agent registry | -| **Brain** | Any model or reasoning substrate | Adapter (roadmap) | +| **Identity** | `agent_ref`, ULID `aura_id`, `ids` trailer | Agent registry + session trailer | +| **Brain** | Any model or reasoning substrate | Integration scripts under `integrations/` | | **Memory** | Any retention backend | Adapter (roadmap) | -| **Tools** | Skills, MCP, HTTP APIs, Skillware | Adapter (roadmap) | -| **Constitution** | Rules, guardrails, constraints on the agent profile | Shipped — constraint engine | -| **Body / Runtime** | The active loop — Python script first | Shipped — runtime helper | -| **Aura** | Harness — hook, enforce, record | Shipped — session + SDK | -| **Audit Trail** | Append-only causal log (`AuraEvent` stream) | Shipped — audit spine / JSONL | -| **Session Export** | Deliverable on close — `.jsonl` + `.summary.json` | Shipped | +| **Tools** | Skills, MCP, HTTP APIs, Skillware | ToolHost + SkillwareHost reference coat | +| **Constitution** | Rules, guardrails, constraints on the agent profile | Constraint engine + manifest merge at bind | +| **Body / Runtime** | The active loop — Python script first | Your host; AURA wraps via SDK or ToolHost | +| **Aura** | Harness — hook, enforce, record | Session + membrane + observers | +| **Audit Trail** | Append-only causal log (`AuraEvent` stream) | Audit spine / JSONL + hash chain | +| **Session Export** | Deliverable on close — JSONL + summary + audit report + OTel | Shipped — see [outputs.md](outputs.md) | --- diff --git a/docs/three-rings.md b/docs/three-rings.md index 346cf20..a9638d9 100644 --- a/docs/three-rings.md +++ b/docs/three-rings.md @@ -1,6 +1,8 @@ # Three Rings -Cross-cutting model — see also [architecture.md](architecture.md). +> **Optional / vision** — cross-cutting ARPA model. For shipped modules see [architecture.md](architecture.md). + +Cross-cutting model — see also [field-services.md](field-services.md). ---