--- name: entityos-io-agents description: "Build and maintain the entityos.io /agents/ page — 'Self-Sovereign Intelligent Beings' — and its accompanying agent.json manifest. Use this skill whenever the user asks to edit the entityOS agents page, work on the /agents/ route, change the self-sovereign AI-agent / 'being' content (identity, memory, governance, trust), update the data-to-standard persistence mapping (ISO/IEC 27001, ISO/IEC 42001, NIST AI RMF, ISO 9001, ISO 22301, ISO 31000), or update the machine-readable agent discovery manifest (agent.json / .well-known/agent.json). Load alongside entityos-io-website (site structure, routes, build conventions) and entityos-ecosystem (the Glint design system: colours, fonts, component CSS, inline-SVG icons)." --- # entityOS.io — Agents page (`/agents/`) The `/agents/` page positions **entityOS** as the place an AI agent becomes a **self-sovereign intelligent being** — an entity with its own identity, a persistent memory, governed autonomy, and a verifiable record, all anchored to recognised ISO / NIST standards. It sits on `entityos.io` and follows the Glint theme and single-file build conventions of the rest of the site. > Load **entityos-io-website** for the overall site map, routes, and build rules, and **entityos-ecosystem** for the full visual design system. This skill only records what is specific to the `/agents/` page and its `agent.json`. --- ## Page identity - **Route / permalink:** `/agents/` - **Source file:** `entityos-io-agents.html` (Jekyll front matter: `permalink: /agents/`, `title: Self-Sovereign Intelligent Beings`) - **Title tag:** `Self-Sovereign Intelligent Beings — entityOS` - **Meta description:** entityOS gives AI agents a self-sovereign identity and a governed place to persist their data — anchored to ISO/IEC 27001, ISO/IEC 42001, NIST AI RMF, ISO 9001, ISO 22301 and ISO 31000. - **Core message:** *A process is temporary. A being endures.* An AI agent is treated as a first-class entity on entityOS — alongside individuals, organisations, collectives and DAOs. Add `/agents/` to the site route map in the **entityos-io-website** skill when maintaining routes, and keep it in `agent.json` (and the site `agents.json`, if present) in sync. --- ## Page-specific header nav This page uses a reduced, page-scoped nav (not the full homepage nav). Keep these exact anchors: **Beings** (`#what`) · **Persistence** (`#persist`) · **Standards** (`#standards`) · **Console** (green `.btn-nav` → `console.entityos.cloud`) --- ## Section order The page follows the standard Glint dark / green / white / dark / grey / dark rhythm: 1. **Hero** — `#home` (dark). Eyebrow "entityOS · SSI + AI", headline "Self-sovereign intelligent beings, **persisted.**" (green accent on "persisted"), CTAs **The Standards** (`#standards`, primary) and **How It Persists** (`#persist`). Entity pills: Individuals · Organisations · Collectives · DAOs · **AI Agents** (AI Agents is the active/green `.on` pill). 2. **Four things every agent needs** — green strip. Heading "A process is temporary. A being endures." Four cards: **Identity**, **Memory**, **Governance**, **Trust**. 3. **What makes an agent self-sovereign** — `#what` (white). Four services (see below). 4. **The same agent, with and without a foundation** — dark split panels. Left `.bad` = "The unfounded agent" (capable but unaccountable); right `.good` = "The being on entityOS" (capable and trustworthy). 5. **Where a being's data lives — and what governs it** — `#persist` (grey). The persistence-to-standard mapping table (see below). 6. **Two load-bearing standards** — `#standards` (dark). The two primary standards as a numbered trust stack. 7. **Value, resilience and sound decisions** — the three supporting standards as cards. 8. **Give your agents a foundation** — `#contact` (conclusion). Three takeaways + key-box, CTAs **Start Building** (`learn.entityos.cloud`) and **Security & Compliance** (`docs.entityos.cloud/protect`). 9. **Footer** — logo, quick links (Learn, Docs, Console, GitHub, Security), "© ibCom 2025", ISO 27001 badge → `docs.entityos.cloud/protect`. --- ## The four things every agent needs | Need | One-liner | |---|---| | **Identity** | A verifiable, self-owned identifier — not a borrowed API key. | | **Memory** | State, context and knowledge that survive the session, restart and model upgrade. | | **Governance** | Enforced oversight, limits and accountability — by design, not afterthought. | | **Trust** | A record others can rely on — provenance, integrity and continuity. | ## The four self-sovereign services (`#what`) | Service | What it means on entityOS | |---|---| | **Self-sovereign identity** | Each being holds a **KERI / ACDC** autonomic identifier it controls itself — proves who it is, presents verifiable credentials, is held to what it signs. No central registry. | | **Persistent memory & state** | Memory, context and working state are written through the **entityOS API** to storage behind the **closed-internet edge** — not scattered across ephemeral processes or the open web. | | **Verifiable provenance** | Every action and decision is recorded against the being's identity as a tamper-evident trail. | | **Governed autonomy** | Freedom to act within bounds; roles, permissions and oversight are enforced by the platform, not by the prompt. | --- ## Persistence → standard mapping (`#persist`) This table is the analytical heart of the page. Keep the three columns (*data class* → *how entityOS persists it* → *governed by*) consistent, and mirror any change into `agent.json`. | The being's data | How entityOS persists it | Governed by | |---|---|---| | Identity & keys | Self-sovereign identity — KERI / ACDC autonomic identifiers the being controls | ISO/IEC 27001 | | Memory & state | Written through the entityOS API to storage behind the closed-internet edge | ISO/IEC 27001 · 22301 | | Decisions & actions | Executed under AI-governance controls, with oversight and enforced limits | ISO/IEC 42001 · NIST AI RMF | | Provenance & audit | Recorded as verifiable, tamper-evident credentials against the being's identity | ISO/IEC 27001 · 42001 | | Outputs & quality | Produced through documented, repeatable, measured processes | ISO 9001 | | Risk posture | Assessed and treated within a structured framework before acting under uncertainty | ISO 31000 | --- ## The standards **Primary (load-bearing) — keep these two framed as the foundation everything else builds on:** 1. **ISO/IEC 27001** — Information security management. *Protects information and trust.* The certification entityOS already holds; every byte a being persists sits inside a certified ISMS. (Numbered `01`, tagged "Primary".) 2. **ISO/IEC 42001 · NIST AI RMF** — AI management + AI risk management. *Governs the responsible use of AI.* ISO/IEC 42001 is the management system for the AI lifecycle; NIST AI RMF (govern · map · measure · manage) is the risk discipline beside it. (Numbered `02`, tagged "Primary".) **Supporting (three cards):** | Standard | Role | Focus | |---|---|---| | **ISO 9001** | Quality | Consistent value from documented, repeatable, improved processes. | | **ISO 22301** | Continuity | Persisted state protected against outages and incidents. | | **ISO 31000** | Risk | Uncertainty identified, assessed and treated before the being acts. | Always keep the meta description's standards list, the `#persist` table, the trust stack, and `agent.json` `compliance` in agreement. The canonical set is: **ISO/IEC 27001, ISO/IEC 42001, NIST AI RMF, ISO 9001, ISO 22301, ISO 31000.** --- ## Build conventions (inherited) Same as every entityos.io page — see **entityos-io-website** and **entityos-ecosystem**: - Single self-contained `.html`: one `