persona-core 0.1.0__tar.gz

persona_core-0.1.0/.gitignore

@@ -0,0 +1,68 @@
+# Python
+__pycache__/
+*.py[cod]
+*.egg-info/
+dist/
+build/
+.venv/
+*.egg
+.mypy_cache/
+.ruff_cache/
+.pytest_cache/
+htmlcov/
+.coverage
+coverage.xml
+
+# Environment
+.env
+.env.local
+.env.*.local
+.secrets/
+
+# IDEs
+.vscode/
+.idea/
+.claude/
+*.swp
+*.swo
+*~
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Node (web app)
+node_modules/
+.next/
+out/
+
+# ChromaDB local storage
+.chroma/
+
+# Persona working directories
+.persona_work/
+.persona_audit/
+
+# UV — uv.lock IS committed (reproducible builds; Dockerfile COPYs it
+# into the image; the CI deploy fails without it).
+
+# rsync atomic-write tempfiles
+.*.??????
+
+# Docker
+docker-compose.override.yml
+
+# Operational caches / temp workspaces
+.playwright-mcp/
+.tmp_smoke_workspace/
+
+# Spec 16 D-16-X-3: inspection artifacts (.docx/.pptx/.xlsx/.pdf) are binary;
+# evidence at close-out is the per-row scorecard in state.md, not the binaries.
+docs/specs/phase2/spec_16/inspection/
+
+# Spec 17 T09 chart-inspection artifacts (.png) are binary; same discipline —
+# scorecards in state.md are the close-out evidence, not the PNG files.
+docs/specs/phase2/spec_17/inspection/
+packages/web/evidence/
+docs/*
+CLAUDE.md

persona_core-0.1.0/CHANGELOG.md

@@ -0,0 +1,48 @@
+# Changelog — persona-core
+
+All notable changes to `persona-core` are recorded here.
+
+The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
+The project follows [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+Per-spec entries are added by the close-out phase of each spec. The authoritative
+project-wide changelog lives at [`/CHANGELOG.md`](../../CHANGELOG.md); this file
+mirrors only the `persona-core`-touching surface.
+
+---
+
+## [Unreleased]
+
+(empty — future post-v0.1 work lands here)
+
+---
+
+## [1.0.0] — 2026-06-07
+
+> **First public, API-stable release of the open-source `persona-core` library under Apache 2.0 (D-11-8).** Includes the Spec V1 JWT-verifier extraction, the Spec 19 amendment set landed during the v0.1 close-out (file-write produced-files / host-out debug logging / credits-service domain relocation), and all prior spec close-outs (Spec 01 schema/stores → Spec 18 routing). Per-package version pin tracks the library's stable public API; product v0.1.0 = the git tag on the system release.
+
+### Added (Spec V1 cross-spec extraction)
+
+- **`persona.auth.jwt_verifier`** — `make_jwt_verifier(config)` + `AuthenticatedUser` factory; new `JwtVerifierConfig: Protocol` (structural subtype satisfied implicitly by `APIConfig` and `VoiceConfig` via the `jwt_algorithms_list` `@property`). `AuthenticationError` relocated from `persona_api.errors` to `persona.errors`; `python-jose[cryptography]` moved to persona-core deps. persona-api re-exports preserve byte-for-byte back-compat (D-V1-X-jwt-verifier-extraction; Spec 08 additive amendment, 9th in the additive-extension chain per D-12-X / D-16-X / D-F4-X-bare-ref-resolution precedent).
+  - Source: [`packages/core/src/persona/auth/jwt_verifier.py`](src/persona/auth/jwt_verifier.py).
+
+### Added (Spec 19 amendment set — chain entries 14 / 15 / 20)
+
+- **L2 (chain 14) D-19-X-file-write-produced-files** — `file_write` tool surfaces produced-files metadata via the produced-files contract so downstream renderers consume host-written outputs identically to sandbox-produced files. Closed-spec additive extension; no Spec 03 reopen.
+- **L4 (chain 15) D-19-X-host-out-debug-logging** — host-out path emits one debug log entry per produced-file emission (loguru `persona.tools.file_write` component). Diagnostic-only; gated behind log level.
+- **L6c (chain 20) D-19-X-credits-service-domain-relocation** — credits-service domain primitives relocated from `persona-api` into `persona-core` (domain home; API keeps the persistence + composition root). Boundary types stay Pydantic v2 frozen `extra="forbid"` per D-12-14 precedent; persona-api `services.credits_service` re-exports the domain shapes to preserve all existing call sites.
+
+### Inherited (close-out roll-up from 2026-05-29 — 2026-06-06)
+
+The following persona-core close-outs are folded into the 1.0.0 anchor; full per-spec rationale lives in [`/CHANGELOG.md`](../../CHANGELOG.md):
+
+- **Spec 17 — Data Analysis and Visualisation.** `data_analysis` built-in skill pack; bytes-persistence layer (`CodeSandbox.copy_produced_file_to` + `read_produced_file_bytes`; `ProducedFileSizeError`); runtime call site for cross-turn `intermediate/*` staging.
+- **Spec 16 — Document Generation Skills.** Four built-in skill packs (`docx_generation`, `pptx_generation`, `xlsx_generation`, `pdf_generation`); M1a runtime affordance (`persona.skills.collect_skill_supplements`); D-16-2-supplements-relative-path production fix.
+- **Spec 15 — Image Generation.** `ImageBackend` Protocol + OpenAI gpt-image-1 + Flux 1.1 [pro] (fal.ai) backends; `identity.visual_style` additive schema extension; three-layer safety with categorical hard-line filter; `make_generate_image_tool` AsyncTool factory; `merge_visual_style` deterministic suffix-conditioning template.
+- **Spec 14 — Document Ingestion.** `DocumentChunk` sibling schema; `DocumentStore` conversation-scoped store; five parsers (txt/md/code, csv, docx, xlsx, pdf) with lazy-import + `MissingDependencyError` discipline; size-aware ingest strategy with D-14-1 3000-token threshold + ladder.
+- **Spec 13 — Vision and Multimodal Input.** `ImageContent` + `ImageRef` schema; `images` JSONB column shape; Pillow downscale-with-hard-ceiling.
+- **Spec 12 — Code Execution Sandbox.** `CodeSandbox` Protocol; `LocalDockerSandbox` (R-12-2 hardening); `make_code_execution_tool`.
+
+### Inherited (prior versions — Spec 11 launch and earlier)
+
+The 1.0.0 anchor subsumes the prior per-spec `[0.x.0]` line: Spec 11 launch (SSRF guard, structured `tool_calls`, streaming `call_id` reconstruction), Spec 04 skills, Spec 03 tools/MCP/Toolbox, Spec 02 backends, Spec 01 foundation (schema, stores, registry, history manager, audit). See [`/CHANGELOG.md`](../../CHANGELOG.md) `[1.0.0]` (2026-05-29) and prior `[0.x.0]` blocks.

persona_core-0.1.0/LICENSE

@@ -0,0 +1,77 @@
+# PolyForm Noncommercial License 1.0.0
+
+<https://polyformproject.org/licenses/noncommercial/1.0.0>
+
+## Acceptance
+
+In order to get any license under these terms, you must agree to them as both strict obligations and conditions to all your licenses.
+
+## Copyright License
+
+The licensor grants you a copyright license for the software to do everything you might do with the software that would otherwise infringe the licensor's copyright in it for any permitted purpose. However, you may only distribute the software according to [Distribution License](#distribution-license) and make changes or new works based on the software according to [Changes and New Works License](#changes-and-new-works-license).
+
+## Distribution License
+
+The licensor grants you an additional copyright license to distribute copies of the software. Your license to distribute covers distributing the software with changes and new works permitted by [Changes and New Works License](#changes-and-new-works-license).
+
+## Notices
+
+You must ensure that anyone who gets a copy of any part of the software from you also gets a copy of these terms, the copyright notice, and the patent notice.
+
+If you modify the software, you must include in any modified copies of the software a prominent notice stating that you have modified the software.
+
+## Changes and New Works License
+
+The licensor grants you an additional copyright license to make changes and new works based on the software for any permitted purpose.
+
+## Patent License
+
+The licensor grants you a patent license for the software that covers patent claims the licensor can license, or becomes able to license, that you would infringe by using the software.
+
+## Noncommercial Purposes
+
+Any noncommercial purpose is a permitted purpose.
+
+## Personal Uses
+
+Personal use for research, experiment, and testing for the benefit of public knowledge, personal study, private entertainment, hobby projects, amateur pursuits, or religious observance, without any anticipated commercial application, is use for a permitted purpose.
+
+## Noncommercial Organizations
+
+Use by any charitable organization, educational institution, public research organization, public safety or health organization, environmental protection organization, or government institution is use for a permitted purpose regardless of the source of funding or obligations resulting from the funding.
+
+## Fair Use
+
+You may have "fair use" rights for the software under the law. These terms do not limit them.
+
+## No Other Rights
+
+These terms do not allow you to sublicense or transfer any of your licenses to anyone else, or prevent the licensor from granting licenses to anyone else. These terms do not imply any other licenses.
+
+## Patent Defense
+
+If you make any written claim that the software infringes or contributes to infringement of any patent, your patent license for the software granted under these terms ends immediately. If your company makes such a claim, your patent license ends immediately for work on behalf of your company.
+
+## Violations
+
+The first time you are notified in writing that you have violated any of these terms, or done anything with the software not covered by your licenses, your licenses can nonetheless continue if you come into full compliance with these terms, and take practical steps to correct past violations, within 32 days of receiving notice. Otherwise, all your licenses end immediately.
+
+## No Liability
+
+***As far as the law allows, the software comes as is, without any warranty or condition, and the licensor will not be liable to you for any damages arising out of these terms or the use or nature of the software, under any kind of legal claim.***
+
+## Definitions
+
+The **licensor** is the individual or entity offering these terms, and the **software** is the software the licensor makes available under these terms.
+
+**You** refers to the individual or entity agreeing to these terms.
+
+**Your company** is any legal entity, sole proprietorship, or other kind of organization that you work for, plus all organizations that have control over, are under the control of, or are under common control with that organization. **Control** means ownership of substantially all the assets of an entity, or the power to direct its management and policies by vote, contract, or otherwise. Control can be direct or indirect.
+
+**Your licenses** are all the licenses granted to you for the software under these terms.
+
+**Use** means anything you do with the software requiring one of your licenses.
+
+---
+
+Copyright 2026 Yasin Hessnawi

persona_core-0.1.0/PKG-INFO

@@ -0,0 +1,197 @@
+Metadata-Version: 2.4
+Name: persona-core
+Version: 0.1.0
+Summary: Build AI personas with typed memory and tier-routed model selection.
+Project-URL: Homepage, https://github.com/yasinhessnawi1/Open-Persona
+Project-URL: Repository, https://github.com/yasinhessnawi1/Open-Persona
+Project-URL: Issues, https://github.com/yasinhessnawi1/Open-Persona/issues
+Project-URL: Changelog, https://github.com/yasinhessnawi1/Open-Persona/blob/main/CHANGELOG.md
+Author-email: Yasin Hessnawi <yasinhessnawi1@gmail.com>
+License-Expression: PolyForm-Noncommercial-1.0.0
+License-File: LICENSE
+Keywords: agent,agentic,ai,llm,persona,pgvector,rag,typed-memory
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: Other/Proprietary License
+Classifier: Operating System :: MacOS
+Classifier: Operating System :: POSIX
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Programming Language :: Python
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Software Development :: Libraries
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Typing :: Typed
+Requires-Python: >=3.11
+Requires-Dist: anthropic<1,>=0.30
+Requires-Dist: chromadb<2,>=1.0
+Requires-Dist: fal-client<2,>=1.0
+Requires-Dist: httpx<1,>=0.27
+Requires-Dist: loguru<1,>=0.7
+Requires-Dist: mcp<2,>=1.0
+Requires-Dist: openai<2,>=1.30
+Requires-Dist: pydantic-settings<3,>=2.3
+Requires-Dist: pydantic<3,>=2.7
+Requires-Dist: python-jose[cryptography]<4,>=3.3
+Requires-Dist: pyyaml<7,>=6.0
+Requires-Dist: regex<2027,>=2024.0
+Requires-Dist: sentence-transformers<4,>=3.0
+Requires-Dist: tiktoken<1,>=0.7
+Requires-Dist: trafilatura<3,>=2.0
+Requires-Dist: typer<1,>=0.12
+Provides-Extra: local
+Requires-Dist: accelerate<1,>=0.31; extra == 'local'
+Requires-Dist: bitsandbytes<1,>=0.43; extra == 'local'
+Requires-Dist: torch<3,>=2.3; extra == 'local'
+Requires-Dist: transformers<5,>=4.42; extra == 'local'
+Provides-Extra: postgres
+Requires-Dist: pgvector<1,>=0.3; extra == 'postgres'
+Requires-Dist: psycopg[binary]<4,>=3.1; extra == 'postgres'
+Requires-Dist: sqlalchemy<3,>=2.0; extra == 'postgres'
+Provides-Extra: sandbox
+Requires-Dist: docker<8,>=7; extra == 'sandbox'
+Provides-Extra: test
+Requires-Dist: pypdf<7,>=6.0; extra == 'test'
+Description-Content-Type: text/markdown
+
+# persona-core
+
+> Source-available Python library for building AI personas with typed memory
+> and tier-routed model selection. Noncommercial use only.
+
+**Status:** PolyForm Noncommercial 1.0.0 · Source Available (Noncommercial Use Only)
+
+## What it is
+
+`persona-core` lets you author a persona as a single YAML document (identity,
+constraints, self-facts, worldview claims with epistemic tags, tools, skills)
+and run it against any of seven model providers — Anthropic, OpenAI, DeepSeek,
+Groq, Together, NVIDIA, OpenRouter, local Ollama, or local HF — with four
+typed memory stores (identity, self-facts, worldview, episodic) backed by
+ChromaDB or Postgres + pgvector. Ships the schema, the four memory stores
+behind a `MemoryStore` protocol, the backend layer behind a `ChatBackend`
+protocol, a sandboxed tool layer (`Toolbox`, MCP client, built-in
+`web_search` / `web_fetch` / `file_read` / `file_write`), a skills layer
+(`SkillScanner` + `SkillInjector` + six built-in skill packs), a vision
+layer (`ImageContent` + `ImageBackend`), document ingestion + generation,
+a code-execution sandbox protocol, an `AuditLogger` protocol with a JSONL
+default, per-component loguru logging, and a `persona` CLI. Every other
+Open Persona package depends on this one; this one depends on nothing inside
+Open Persona.
+
+## Install
+
+```bash
+pip install persona-core                 # core + Chroma + frontier SDKs
+pip install persona-core[local]          # adds torch / transformers for local HF inference
+pip install persona-core[postgres]       # adds psycopg + pgvector for the Postgres backend
+pip install persona-core[sandbox]        # adds docker SDK for LocalDockerSandbox
+```
+
+Python ≥ 3.11.
+
+For workspace development from the monorepo:
+
+```bash
+git clone https://github.com/yasinhessnawi1/Open-Persona.git
+cd open-persona
+uv sync --all-packages
+```
+
+## Run
+
+Author and chat with a persona from the terminal:
+
+```bash
+persona init                                      # interactive → astrid.yaml
+persona validate examples/astrid_tenancy_law.yaml
+export PERSONA_PROVIDER=deepseek
+export PERSONA_MODEL=deepseek-chat
+export PERSONA_API_KEY=<your-key>
+persona chat examples/astrid_tenancy_law.yaml
+persona audit examples/astrid_tenancy_law.yaml    # tail the JSONL audit log
+```
+
+Three example personas ship in [`examples/`](examples/):
+`astrid_tenancy_law.yaml` (Norwegian tenancy-law assistant),
+`kai_research.yaml` (research assistant), `maren_writing_coach.yaml`
+(tool-free writing coach).
+
+Programmatic use:
+
+```python
+import asyncio
+from pathlib import Path
+
+from persona.schema.persona import Persona
+from persona.backends import OpenAICompatibleBackend, BackendConfig
+from persona.schema.conversation import ConversationMessage
+
+async def main() -> None:
+    persona = Persona.from_yaml(Path("examples/astrid_tenancy_law.yaml"))
+    backend = OpenAICompatibleBackend(
+        BackendConfig(provider="deepseek", model="deepseek-chat")
+    )
+    system = f"You are {persona.identity.name}, {persona.identity.role}."
+    reply = await backend.chat([
+        ConversationMessage(role="system", content=system, created_at=None),
+        ConversationMessage(role="user", content="Hva sier husleieloven om mugg?", created_at=None),
+    ])
+    print(reply.content)
+
+asyncio.run(main())
+```
+
+For the full conversation loop with router, tool dispatch, episodic
+write-back and per-turn logging, compose with `persona-runtime`.
+
+## Test
+
+```bash
+uv run pytest packages/core                          # unit + contract (default)
+uv run pytest packages/core -m integration           # needs Postgres in Docker
+uv run mypy packages/core/src --strict
+uv run ruff check packages/core
+```
+
+## Highlights
+
+- Frozen Pydantic v2 boundary models, `extra="forbid"` everywhere
+- Deterministic chunk IDs: `{persona_id}::{store_kind}::{index:04d}`
+- Versioned append-only stores; identity is immutable at runtime
+- SHA-256 `content_hash` on every chunk; one `AuditEvent` per mutation
+- Three-source write policy: `system` / `user` / `persona_self`
+- 2k-token-budgeted skill injection (`SkillInjector.TOKEN_BUDGET`)
+- Native tool calls (Anthropic / OpenAI / DeepSeek / Groq / Together / NVIDIA /
+  OpenRouter) + prompt-shim fallback (Ollama / HF local)
+- MCP Streamable HTTP client + adapter
+- Sandboxed file tools (path resolver rejects `..`, abs paths, symlink
+  escape, NUL bytes, mixed separators)
+- Six built-in skill packs: `web_research`, `document_drafting`,
+  `docx_generation`, `pptx_generation`, `xlsx_generation`, `pdf_generation`
+- Image generation backends (OpenAI gpt-image-1, fal.ai Flux 1.1 [pro])
+  with three-layer safety + categorical hard-line filter
+- Vision input (`ImageContent`) + document ingestion + a `CodeSandbox`
+  protocol with `LocalDockerSandbox` reference implementation
+
+## Architecture role
+
+`persona-core` is layer 4 of the Open Persona stack — the source-available
+foundation. A persona is a YAML document; the schema, the typed memory
+stores, the model-provider adapters, and the tool/skill machinery all live
+here. `persona-runtime` composes the orchestration loop on top of this
+library; `persona-api` exposes it over HTTP; `persona-web` is the browser
+front-end. The dependency arrow points one way — `persona-core` imports
+nothing from the upper layers.
+
+## Contribute
+
+Contributions welcome under the same PolyForm Noncommercial 1.0.0 license.
+The package is source-available for noncommercial use; commercial use
+requires a separate license — contact the rights holder. Issues and pull
+requests welcome at
+[github.com/yasinhessnawi1/Open-Persona](https://github.com/yasinhessnawi1/Open-Persona).
+See [LICENSE](LICENSE) and the package
+[`SPEC.md`](SPEC.md) for the public surface.

persona_core-0.1.0/README.md

@@ -0,0 +1,139 @@
+# persona-core
+
+> Source-available Python library for building AI personas with typed memory
+> and tier-routed model selection. Noncommercial use only.
+
+**Status:** PolyForm Noncommercial 1.0.0 · Source Available (Noncommercial Use Only)
+
+## What it is
+
+`persona-core` lets you author a persona as a single YAML document (identity,
+constraints, self-facts, worldview claims with epistemic tags, tools, skills)
+and run it against any of seven model providers — Anthropic, OpenAI, DeepSeek,
+Groq, Together, NVIDIA, OpenRouter, local Ollama, or local HF — with four
+typed memory stores (identity, self-facts, worldview, episodic) backed by
+ChromaDB or Postgres + pgvector. Ships the schema, the four memory stores
+behind a `MemoryStore` protocol, the backend layer behind a `ChatBackend`
+protocol, a sandboxed tool layer (`Toolbox`, MCP client, built-in
+`web_search` / `web_fetch` / `file_read` / `file_write`), a skills layer
+(`SkillScanner` + `SkillInjector` + six built-in skill packs), a vision
+layer (`ImageContent` + `ImageBackend`), document ingestion + generation,
+a code-execution sandbox protocol, an `AuditLogger` protocol with a JSONL
+default, per-component loguru logging, and a `persona` CLI. Every other
+Open Persona package depends on this one; this one depends on nothing inside
+Open Persona.
+
+## Install
+
+```bash
+pip install persona-core                 # core + Chroma + frontier SDKs
+pip install persona-core[local]          # adds torch / transformers for local HF inference
+pip install persona-core[postgres]       # adds psycopg + pgvector for the Postgres backend
+pip install persona-core[sandbox]        # adds docker SDK for LocalDockerSandbox
+```
+
+Python ≥ 3.11.
+
+For workspace development from the monorepo:
+
+```bash
+git clone https://github.com/yasinhessnawi1/Open-Persona.git
+cd open-persona
+uv sync --all-packages
+```
+
+## Run
+
+Author and chat with a persona from the terminal:
+
+```bash
+persona init                                      # interactive → astrid.yaml
+persona validate examples/astrid_tenancy_law.yaml
+export PERSONA_PROVIDER=deepseek
+export PERSONA_MODEL=deepseek-chat
+export PERSONA_API_KEY=<your-key>
+persona chat examples/astrid_tenancy_law.yaml
+persona audit examples/astrid_tenancy_law.yaml    # tail the JSONL audit log
+```
+
+Three example personas ship in [`examples/`](examples/):
+`astrid_tenancy_law.yaml` (Norwegian tenancy-law assistant),
+`kai_research.yaml` (research assistant), `maren_writing_coach.yaml`
+(tool-free writing coach).
+
+Programmatic use:
+
+```python
+import asyncio
+from pathlib import Path
+
+from persona.schema.persona import Persona
+from persona.backends import OpenAICompatibleBackend, BackendConfig
+from persona.schema.conversation import ConversationMessage
+
+async def main() -> None:
+    persona = Persona.from_yaml(Path("examples/astrid_tenancy_law.yaml"))
+    backend = OpenAICompatibleBackend(
+        BackendConfig(provider="deepseek", model="deepseek-chat")
+    )
+    system = f"You are {persona.identity.name}, {persona.identity.role}."
+    reply = await backend.chat([
+        ConversationMessage(role="system", content=system, created_at=None),
+        ConversationMessage(role="user", content="Hva sier husleieloven om mugg?", created_at=None),
+    ])
+    print(reply.content)
+
+asyncio.run(main())
+```
+
+For the full conversation loop with router, tool dispatch, episodic
+write-back and per-turn logging, compose with `persona-runtime`.
+
+## Test
+
+```bash
+uv run pytest packages/core                          # unit + contract (default)
+uv run pytest packages/core -m integration           # needs Postgres in Docker
+uv run mypy packages/core/src --strict
+uv run ruff check packages/core
+```
+
+## Highlights
+
+- Frozen Pydantic v2 boundary models, `extra="forbid"` everywhere
+- Deterministic chunk IDs: `{persona_id}::{store_kind}::{index:04d}`
+- Versioned append-only stores; identity is immutable at runtime
+- SHA-256 `content_hash` on every chunk; one `AuditEvent` per mutation
+- Three-source write policy: `system` / `user` / `persona_self`
+- 2k-token-budgeted skill injection (`SkillInjector.TOKEN_BUDGET`)
+- Native tool calls (Anthropic / OpenAI / DeepSeek / Groq / Together / NVIDIA /
+  OpenRouter) + prompt-shim fallback (Ollama / HF local)
+- MCP Streamable HTTP client + adapter
+- Sandboxed file tools (path resolver rejects `..`, abs paths, symlink
+  escape, NUL bytes, mixed separators)
+- Six built-in skill packs: `web_research`, `document_drafting`,
+  `docx_generation`, `pptx_generation`, `xlsx_generation`, `pdf_generation`
+- Image generation backends (OpenAI gpt-image-1, fal.ai Flux 1.1 [pro])
+  with three-layer safety + categorical hard-line filter
+- Vision input (`ImageContent`) + document ingestion + a `CodeSandbox`
+  protocol with `LocalDockerSandbox` reference implementation
+
+## Architecture role
+
+`persona-core` is layer 4 of the Open Persona stack — the source-available
+foundation. A persona is a YAML document; the schema, the typed memory
+stores, the model-provider adapters, and the tool/skill machinery all live
+here. `persona-runtime` composes the orchestration loop on top of this
+library; `persona-api` exposes it over HTTP; `persona-web` is the browser
+front-end. The dependency arrow points one way — `persona-core` imports
+nothing from the upper layers.
+
+## Contribute
+
+Contributions welcome under the same PolyForm Noncommercial 1.0.0 license.
+The package is source-available for noncommercial use; commercial use
+requires a separate license — contact the rights holder. Issues and pull
+requests welcome at
+[github.com/yasinhessnawi1/Open-Persona](https://github.com/yasinhessnawi1/Open-Persona).
+See [LICENSE](LICENSE) and the package
+[`SPEC.md`](SPEC.md) for the public surface.