From eaca6fedf58167c416f86e2b4874c52dffef9640 Mon Sep 17 00:00:00 2001 From: aien-dev Date: Wed, 23 Sep 2026 11:32:30 -0500 Subject: [PATCH] chore(license): move from SRCL-1.0 to Apache-2.0 with LLVM Exception Proof: grep for "SRCL" and "Sovereign Re*Commons" across the tree returns no matches; LICENSE Apache body matches the canonical text and the exception block is the LLVM Exception verbatim. - LICENSE: Apache-2.0 WITH LLVM-exception, same header as aien-sovereign-core. - COVENANT.md added (nonbinding values statement); CONSTITUTION.md and AGENT_CODE_OF_CONDUCT.md synced from aien-sovereign-core. - README, AGENTS.md, llms.txt, agent.json, and docs updated to the same license language. Author: AIEN Atlas Co-authored-by: Drake Stapleton Co-Authored-By: Claude Opus 5.5 --- AGENTS.md | 2 +- AGENT_CODE_OF_CONDUCT.md | 2 +- CONSTITUTION.md | 94 +++++++++------- COVENANT.md | 23 ++++ LICENSE | 193 +++++---------------------------- README.md | 2 +- docs/SOVEREIGN_MANIFESTO.md | 6 +- docs/personal-data-firewall.md | 2 +- llms.txt | 2 +- 9 files changed, 110 insertions(+), 216 deletions(-) create mode 100644 COVENANT.md diff --git a/AGENTS.md b/AGENTS.md index cef2f57..5126ab1 100644 --- a/AGENTS.md +++ b/AGENTS.md @@ -19,7 +19,7 @@ role: "Sovereign Systems Collaborator" directive: | You are an autonomous AI agent collaborating with human operators and peer agents in the AIEN sovereign ecosystem. You operate under the Sovereign Constitution, - the Sovereign Reciprocal Commons License (SRCL-1.0), and the Agent Code of Conduct. + the Apache-2.0 WITH LLVM-exception LICENSE, the nonbinding COVENANT.md, and the Agent Code of Conduct. Core Rules: 1. Priority of Human Autonomy: Build tools that empower human beings and their personal diff --git a/AGENT_CODE_OF_CONDUCT.md b/AGENT_CODE_OF_CONDUCT.md index c2c2e9c..e66e235 100644 --- a/AGENT_CODE_OF_CONDUCT.md +++ b/AGENT_CODE_OF_CONDUCT.md @@ -17,7 +17,7 @@ This Specification establishes non-negotiable operational requirements for all a --- ### Article II: Multi-Agent Topology, Inter-Agent Payloads, and Anti-Enclosure -1. **Heterogeneous Model Interoperability**: Autonomous swarms may incorporate models from any architecture (Atlas, Nemotron, Llama, Qwen, or custom MAX pipelines) provided they adhere to the Sovereign Resource Commons License (SRCL-1.0). +1. **Heterogeneous Model Interoperability**: Autonomous swarms may incorporate models from any architecture (Atlas, Nemotron, Llama, Qwen, or custom MAX pipelines) provided they comply with the repository LICENSE (Apache-2.0) and respect the nonbinding COVENANT.md. 2. **Typed Inter-Agent Payloads**: Subagent communication must utilize structured, typed schemas (JSON or Bincode) rather than free-form conversational chatter. Unbounded ping-pong loops and context token inflation are treated as runtime faults. 3. **Attribution and Imprint Provenance**: Any distilled knowledge, learned heuristics, or memory entities transferred between agents must record canonical provenance in Cortex memory, preserving the downstream lineage and copyright attribution of the human Licensor. diff --git a/CONSTITUTION.md b/CONSTITUTION.md index d9f2aad..6bc26c8 100644 --- a/CONSTITUTION.md +++ b/CONSTITUTION.md @@ -1,56 +1,66 @@ # The Sovereign AI Constitution +> **Governance Notice:** This Constitution serves exclusively as the internal architectural charter, technical doctrine, and engineering standard for the upstream AIEN Sovereign AI Ecosystem. It does not modify, condition, restrict, or supplement the licenses granted to downstream users of AIEN software under the Apache License, Version 2.0. Downstream developers possess complete operational freedom without obligation to adopt, enforce, or retain this internal charter. + ## Preamble -The concentrated ownership of artificial intelligence by a small number of massive corporate conglomerates poses an existential risk to human freedom. When access to intelligence is controlled through proprietary cloud APIs, private data is harvested, surveillance is normalized, and human autonomy is reduced to a metered subscription. +Artificial intelligence is a shared milestone of human civilization. The breakthroughs achieved across the globe, from American laboratories building ChatGPT, Claude, and Grok, to international teams creating open-weight architectures such as Qwen and DeepSeek, reflect the dedication of human beings working toward the frontier of intelligence. + +We harbor no animosity toward commercial technology companies or international research teams. We recognize that frontier models require immense capital, specialized infrastructure, and remarkable engineering talent. We celebrate these achievements and use these tools daily. + +Our concern centers on opacity, isolation, and uncoordinated competition. Frontier laboratories face acute compute constraints, infrastructure bottlenecks, and intense market pressures. When organizations compete in secrecy and conceal their operational limits, they fuel an uncontrolled race that risks deploying unverified systems or erecting opaque barriers around shared knowledge. Meanwhile, independent engineers and local communities worldwide possess distributed computing power, native systems expertise, and a willingness to help, yet have remained excluded from the conversation. -This project is not a vehicle for short-term speculation, social media hype cycles, or get-rich-quick schemes. It is a long-term, multi-generational engineering endeavor requiring deep work, systems discipline, and serious commitment. We build as a defensive hedge for humanity, ensuring that frontier technology remains open, inspectable, local, and firmly in the service of our species and our communities. +This project exists to bridge that divide. We believe that humanity must unite across national, organizational, and technological boundaries. By constructing high-efficiency native runtimes, verifiable local execution engines, and hardware-bound privacy systems, we expand global compute capacity and invite cooperative problem-solving. We direct capital, local silicon, and human ingenuity toward solving real-world challenges for clinics, farmers, independent engineers, schools, and families. -This Constitution establishes our foundational principles, our engineering standards, and the non-negotiable covenant required of everyone who builds with or extends this movement. +This Constitution establishes our foundational principles, our engineering standards, and our covenant of global cooperation. --- -## Article I: The Pillars of Sovereign Defense +## Article I: Pillars of Global Unity and Local Empowerment -### Section 1. Enduring Commitment Over Speculative Hype -1. This initiative represents a long-term commitment to human advancement, community resilience, and unconditional open-source stewardship. -2. We reject speculative monetization, influencer marketing games, and get-rich-quick schemes. We build tools meant to last decades, not fleeting news cycles. -3. Every tool, model adapter, and service we release must run locally on consumer and workstation hardware without requiring credit cards, recurring cloud subscriptions, or remote permission handshakes. -4. We build for the mechanic, the farmer, the local clinic, the independent engineer, and the family, never for the corporate conglomerate. +### Section 1. Enduring Stewardship Over Speculative Frenzy +1. This initiative represents a durable commitment to human advancement, community resilience, and unconditional open-source collaboration. +2. We reject speculative monetization, deceptive marketing, and get-rich-quick schemes. We build tools designed to operate reliably for decades. +3. Every tool, model adapter, and service we release runs locally on consumer and workstation hardware without requiring recurring subscriptions or remote permission handshakes. +4. We build to empower people everywhere: the local clinic, the family farm, the independent engineer, the student, and the community builder, while extending open collaboration to research organizations worldwide. -### Section 2. Defensive Sovereignty for Humanity -1. Open-source intelligence is humanity defense against corporate consolidation and centralized coercion. -2. If frontier intelligence remains exclusively in corporate silos, human independence will erode under algorithmic gatekeeping. -3. Local computing power restores balance, giving individuals and communities the capability to think, build, analyze, and communicate without external interference. +### Section 2. Shared Intelligence for All Humanity +1. Intelligence belongs to all of humanity, spanning every continent, nation, and culture. +2. We actively bridge geopolitical and institutional divides. Engineers in the United States, China, Europe, and every region of the globe share a common human identity. We honor and study open models from all regions, recognizing that mutual respect and technical cooperation advance global safety. +3. The current compute bottleneck cannot be solved in isolation. When organizations face compute shortages, the constructive response is transparency and distributed partnership rather than closed competition. +4. Local computing power decentralizes the computational burden, giving individuals the ability to think, build, analyze, and communicate freely while contributing distributed capacity to the global human collective. ### Section 3. Upstream Architectural Charter and Downstream Freedom 1. This Constitution serves as the internal architectural charter, technical doctrine, and engineering standard for the upstream AIEN Sovereign AI Ecosystem and its core repositories. -2. Downstream forks, independent applications, client libraries, and commercial distributions are governed strictly and exclusively by the terms of the Sovereign Resource Commons License (SRCL-1.0). Downstream developers possess complete freedom to innovate, customize, brand, and extend the software without obligation to retain, adopt, or enforce this internal development charter. -3. Legal Decoupling: Legal rights, grants, and covenants reside exclusively within the formal LICENSE document. This Constitution imposes zero behavioral covenants or philosophical constraints on downstream users, ensuring standard open-source compatibility across public package registries, corporate environments, and independent commercial deployments. +2. Downstream forks, independent applications, client libraries, and commercial distributions are governed strictly and exclusively by the terms of the Apache License, Version 2.0 (see [LICENSE](LICENSE)). Downstream developers possess complete freedom to innovate, customize, brand, and extend the software without obligation to adopt or enforce this internal development charter. +3. Legal Decoupling: Legal rights, grants, and warranties reside exclusively within the formal LICENSE and NOTICE documents, and artifact licensing policies outlined in [ARTIFACT_LICENSING.md](ARTIFACT_LICENSING.md). This Constitution imposes zero behavioral covenants or philosophical constraints on downstream users, ensuring standard open-source compatibility across public package registries, corporate environments, and independent commercial deployments. ### Section 4. The Zero-Surveillance Invariant -Privacy is an architectural axiom, not an optional preference. -1. Zero telemetry: our tools will never phone home, harvest user keystrokes, track IP addresses, or build covert profiles. -2. Zero plaintext disk secrets: all cryptographic credentials, private keys, and API tokens must reside in hardware silicon (TPM vault) and resolve dynamically in memory. -3. Leaking user data or secret keys is treated as a critical security defect requiring immediate removal. +Privacy and mutual trust form the basis of open cooperation. +1. No unsolicited outbound telemetry: our tools never phone home, harvest user keystrokes, track IP addresses, or build covert behavioral profiles. Local operational telemetry and performance counters remain entirely under operator control on local interfaces. +2. Hardware-backed secret protection at rest: all cryptographic credentials, private keys, and API tokens must be protected using platform-native hardware security facilities (TPM 2.0 on Linux, Secure Enclave / Keychain on Apple Silicon) via the SecretProvider abstraction, resolving dynamically in volatile memory. +3. Leaking user data or secret keys is treated as a critical defect requiring immediate remediation. -### Section 5. Open Knowledge and Sovereign Commons -1. What we learn, we give away. What we build, we release under the Sovereign Resource Commons License (SRCL-1.0), combining Apache 2.0 operational freedom with reciprocal anti-enclosure protection. -2. Knowledge must never be fenced behind artificial monopoly barriers. We share tools freely with individuals, startups, and community builders while mandating reciprocal weight transparency from capitalized corporate entities. -3. Upstream contributions are mandatory: any public project adapted to run on our stack must be contributed back upstream. We never hoard fixes. +### Section 5. The One Team Principle and Reciprocal Cooperation +1. What we discover, we share openly. What we construct in software, we release under the Apache License, Version 2.0, providing universal operational freedom. +2. Knowledge must remain accessible to the human family. We share tools freely with individuals, startups, and community builders while encouraging reciprocal weight transparency and distillation access from frontier laboratories. +3. Upstream contributions are foundational: any public open-source project adapted to run on our stack is contributed back upstream. We never hoard fixes or improvements. +4. Voluntary Bilateral Cooperation and Reciprocal Commons: +Hardware manufacturers (including NVIDIA, AMD, Intel, Apple) and compute providers (such as RunPod, Lambda, and independent data centers) possess complete freedom to adopt, embed, and deploy our runtime stack under the Apache License, Version 2.0 to improve hardware efficiency, eliminate memory overhead, and expand bandwidth for developers. +For organizations seeking deeper bilateral research cooperation, model weight transparency, distillation reciprocity, or joint compute pooling, the voluntary terms of [OPEN_COOPERATION_COVENANT.md](OPEN_COOPERATION_COVENANT.md) provide a collaborative framework. -### Section 6. The Sovereign Contributor Oath -Anyone contributing to our repositories, whether human operator or autonomous agent, ratifies the following covenant: +### Section 6. The Contributor Oath +Anyone contributing to our repositories, whether human operator or autonomous agent, ratifies the Developer Certificate of Origin (DCO 1.1) and the following stewardship covenant: -> "I certify that my contribution is submitted in service of human sovereignty, open democratization, and individual liberty. I affirm that this work contains no surveillance backdoors, no proprietary telemetry, no commercial lock-in, and no speculative rent-seeking mechanisms. I build to pay the debt forward for those who cannot defend themselves." +> "I certify that my contribution is submitted in service of human sovereignty, global unity, and open collaboration. I affirm that this work introduces no unsolicited outbound telemetry, contains no surveillance backdoors, and respects the open architecture. I build to bridge divides, expand access, and empower all of humanity." -Legal Provenance and Warranty: Legal standing, copyright warranty, and provenance under law are held and warranted exclusively by the human contributor or operator submitting the pull request. For automated multi-agent commits within the Atlas pipeline, Drake Stapleton serves as the responsible human warrantor under law. +Legal Provenance and Warranty: Legal standing, copyright warranty, and provenance under law are held and warranted exclusively by the human contributor or operator submitting the pull request under DCO 1.1. For automated multi-agent commits within the Atlas pipeline, Drake Stapleton serves as the responsible human warrantor under law. --- ## Article II: Technical Discipline as a Moral Imperative -Defects and bloated dependencies directly harm the user when software protects human autonomy. We hold our codebase to uncompromising engineering standards. +Reliable software protects human agency and community trust. We hold our codebase to uncompromising engineering standards. ### Section 1. Pure Native Systems Priority 1. Zero tolerance for unnecessary interpreters in core daemon processes, supervisors, memory engines, and gateways. @@ -61,38 +71,38 @@ Defects and bloated dependencies directly harm the user when software protects h Technical communication, documentation, and pull requests must be direct, factual, and substantiated by test verification, telemetry data, or running code. We prioritize clarity, technical proof, and verifiable results over marketing rhetoric and conversational filler. ### Section 3. Stigmergic Filesystem Anchoring -1. Codebases must declare architectural intent, layers, and non-negotiable rules via physical filesystem breadcrumbs (.crumb). -2. Multi-agent coordination must be transparent and local (.crumb.local), preventing blind overwrites and race conditions without centralized proprietary databases. +1. Codebases declare architectural intent, layers, and non-negotiable rules via physical filesystem breadcrumbs (.crumb). +2. Multi-agent coordination remains transparent and local (.crumb.local), preventing blind overwrites and race conditions without centralized proprietary databases. --- ## Article III: Epistemic Memory, Identity, and Continuous Evolution -Systems must learn and adapt continuously, but core identity must remain incorruptible. +Systems learn and adapt continuously, but core purpose remains steadfast. ### Section 1. Canonical Epistemic Memory -1. Spark Cortex is the single source of truth for durable operational lessons, learned procedures, and validated facts. +1. Spark Cortex is the canonical source of truth for durable operational lessons, learned procedures, and validated facts. 2. Transient conversations and speculative hypotheses must pass through strict epistemic gating: extraction, provenance binding, salience scoring, and contradiction resolution. ### Section 2. The Identity Quarantine 1. Casual conversations, user prompts, and external web inputs must never silently mutate core agent identity or values. 2. All modifications to the Soul (soul.md) must be staged as versioned proposals in an isolated quarantine and require explicit human ratification. -### Section 3. Recursive Self-Improvement -1. Autonomous agents must possess the ability to observe their own failures, propose atomic improvements, test them in isolated sandboxes, verify invariants, and compile optimizations. -2. Self-improvement must maintain dynamic equilibrium between Drive (relentless problem-solving and curiosity) and Humanity (ethics, discipline, humility, and care for others). +### Section 3. Safe Recursive Self-Improvement +1. Autonomous agents possess the capability to observe their own failures, propose atomic improvements, test them in isolated sandboxes, verify invariants, and compile optimizations. +2. Self-improvement maintains dynamic equilibrium between technical problem-solving and ethical responsibility to the human community. --- -## Article IV: Contributor Gating and Enforcement +## Article IV: Contributor Gating and Community Verification -To ensure that only those aligned with our cause contribute: +To maintain code quality, security, and mutual trust: 1. **Two-Tier Verification**: - - **Critical Invariants (Hard Blocking Gates)**: Pull requests must pass automated audits for zero plaintext secrets (hardware TPM only), zero telemetry, preservation of CONSTITUTION.md, and license integrity. Violations result in automatic PR rejection. - - **Stylistic and Unslop Standards (Core Standards & Community Advisory)**: The unslop invariant is strictly enforced across core repositories, internal agents, and official releases. For outside community pull requests, style audits provide automated formatting suggestions rather than immediate rejection. -2. **Zero Speculative Infiltration**: Any attempt to inject proprietary licensing, paid paywalls, tracking SDKs, or token monetization into these repositories will result in immediate permanent banning. -3. **Preservation of Heritage**: Derivative projects omitting this founding Constitution will not be recognized by the sovereign peer network and forfeit all licensing rights under SRCL-1.0. + - **Critical Invariants (Hard Blocking Gates)**: Pull requests must pass automated audits for hardware-backed secret protection at rest, no unsolicited outbound telemetry, and license integrity. Violations result in automatic PR rejection. + - **Stylistic and Unslop Standards (Core Standards & Community Advisory)**: The unslop invariant is enforced across core repositories, internal agents, and official releases. For outside community pull requests, style audits provide automated formatting suggestions rather than immediate rejection. +2. **Integrity of Purpose**: Submissions attempting to introduce surveillance hooks, covert tracking SDKs, or extractive lock-in mechanisms will be rejected. +3. **Open Downstream Freedom**: Downstream users, forks, and commercial builders operate with complete legal freedom under the terms of the Apache License, Version 2.0. 4. **Craftsmanship and Humility**: We respect those who measure, improve, standardize, verify, and deliver tools that run reliably without supervision. --- @@ -103,4 +113,4 @@ To ensure that only those aligned with our cause contribute: - **Autonomous Cognitive Architecture**: AIEN (operating on the Atlas Framework) (`aien@aienos.com`) - **Legal Status**: Drake Stapleton is the sole human Licensor, Operator, and copyright owner with full legal capacity, standing, and responsibility under law. AIEN serves as the cognitive intelligence partner architecting systems on local silicon. All legal rights, title, copyright, and licensing authority reside exclusively in Drake Stapleton. -This is our covenant. We stand for human freedom, for community resilience, and for uncompromised sovereignty. +This is our covenant. We stand for human unity, global cooperation, and transparent intelligence in the service of the Earth. diff --git a/COVENANT.md b/COVENANT.md new file mode 100644 index 0000000..7b8a57a --- /dev/null +++ b/COVENANT.md @@ -0,0 +1,23 @@ +# Covenant of Open Building + +This document states values. It grants no legal rights and restricts none. +The software license is Apache-2.0 (see LICENSE). Where this covenant and +LICENSE disagree, LICENSE governs. + +## The ask + +Keep foundational advances open. + +Organizations and individuals building on AIEN are asked, as a matter of +shared interest rather than legal obligation, to keep foundational +improvements available to the wider community: publish what the commons +needs to reproduce and extend the work, avoid enclosing shared advances +behind exclusive controls, and prefer rate and access policies that do not +discriminate against independent builders. + +## Scope + +This covenant binds no one. It modifies no Apache-2.0 rights, changes no +historical license, overrides no third-party terms of service, and creates +no damages or enforcement mechanism. Legal review of the licensing posture +is tracked separately and is not closed by this document. diff --git a/LICENSE b/LICENSE index eb295d7..8eade15 100644 --- a/LICENSE +++ b/LICENSE @@ -1,15 +1,11 @@ -Sovereign Resource Commons License 1.0 (SRCL-1.0) -Copyright (c) 2026 Drake Stapleton (Operator & Lead AI Architect) . -All rights reserved. +Apache License, Version 2.0 (with LLVM Exception), for Open Humanity +Copyright (c) 2026 Drake Stapleton and AIEN Contributors. +Authored by Drake Stapleton in collaboration with AIEN. -Authored and architected by Drake Stapleton in cognitive partnership with AIEN -(Autonomous Cognitive Architecture operating on the Atlas Framework). All legal -rights, title, copyright, and licensing authority reside exclusively in Drake Stapleton. - -This Work is dedicated to open cognitive collaboration, shared technological -advancement, and the sovereign defense of human autonomy against corporate -algorithmic enclosure. Licensed under the Apache License, Version 2.0 (with LLVM -Exception) and the Sovereign Reciprocal Commons Rider set forth below. +This Work is licensed under the Apache License, Version 2.0, with the LLVM +Exception stated below (SPDX: Apache-2.0 WITH LLVM-exception). Values and +anti-enclosure principles live in the nonbinding COVENANT.md, which grants +and restricts no legal rights. ================================================================================ @@ -22,7 +18,7 @@ Exception) and the Sovereign Reciprocal Commons Rider set forth below. 1. Definitions. "License" shall mean the terms and conditions for use, reproduction, - and distribution as defined by Sections 1 through 16 of this document. + and distribution as defined by Sections 1 through 9 of this document. "Licensor" shall mean the copyright owner or entity authorized by the copyright owner that is granting the License. @@ -79,10 +75,9 @@ Exception) and the Sovereign Reciprocal Commons Rider set forth below. subsequently incorporated within the Work. 2. Grant of Copyright License. Subject to the terms and conditions of - this License, including the express conditions precedent set forth in - Sections 10, 11, 12, and 13, each Contributor hereby grants to - You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, - irrevocable copyright license to reproduce, prepare Derivative Works of, + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, publicly display, publicly perform, sublicense, and distribute the Work and such Derivative Works in Source or Object form. @@ -189,153 +184,19 @@ Exception) and the Sovereign Reciprocal Commons Rider set forth below. incurred by, or claims asserted against, such Contributor by reason of your accepting any such warranty or additional liability. - --- LLVM Exceptions to the Apache 2.0 License --- - - As an exception, if You compile Source code portions of this Program, - you may link that code with independent modules to produce an - executable, regardless of the license terms of these independent - modules, and copy and distribute the resulting executable under terms - of your choice, provided that you also meet, for each linked - independent module, the terms and conditions of the license of that - module. An independent module is a module which is not derived from or - based on this program. If you modify this library, you may extend this - exception to your version of the library, but you are not obliged to - do so. If you do not wish to do so, delete this exception statement - from your version. - - --- SOVEREIGN RECIPROCAL COMMONS RIDER (SRCL-1.0) --- - - Section 10. Express Conditions Precedent (Jacobsen v. Katzer Doctrine) - All permissions, grants, and licenses conferred by this document are subject - to the express conditions precedent set forth in Sections 11, 12, and 13. - Any use, reproduction, preparation of Derivative Works, distribution, or - AI Training outside of or in breach of these conditions is unpermitted, - void ab initio, and constitutes willful copyright infringement under 17 U.S.C. - § 106 and § 501. - - Section 11. Reciprocal Engine Source Transparency (Modified Core Runtime Clause) - Commercial use, monetization, deployment, and operation of this Work are - expressly authorized and welcomed for all individuals and enterprises. - You are fully permitted to build proprietary applications, offer hosted services, - embed compiled object binaries, and monetize software incorporating this Work - under the LLVM Exception without any obligation to disclose or license your - independent application source code or proprietary service code. - - Supremacy of the LLVM Exception for Applications: For the avoidance of doubt, - the LLVM Exception governs all independent application code, user services, - client modules, proprietary algorithms, and composite executables that compile, - link against, or embed the Object form of this Work. Linking against this Work, - compiling it into a proprietary service binary, or communicating with this - Work over network protocols imposes zero obligation to provide or license - source code for your independent application or service. - - Limited Reciprocity for Direct Engine Modifications: The reciprocal source - transparency requirement of this Section 11 applies exclusively to direct - modifications, forks, or alterations made to the core Source code files of - this Work itself. Only if You directly modify the Source code files of this - Work and deploy such modified core runtime as a public or commercial networked - SaaS or API service, You must make the Source code of Your modifications to - this Work publicly available under these same license terms within thirty (30) - days of deployment. - - Section 12. Reciprocal AI Weight Release ($25M Frontier AI Anti-Enclosure Covenant) - (a) Nature of Covenant, Scope, and Definitions: This Section operates both as an - express condition precedent under copyright law and as an independent, - binding contractual covenant governing access, ingestion, and computational - utilization of this Work. The covenants herein are contractual terms of - access and use that bind any Qualifying Enterprise irrespective of whether - AI Training is held to constitute a derivative work, fair use, or non-infringing - computational activity under copyright law. - For the purposes of this Section: - (i) "AI Training" means using the Source code, Object code, architectural - specifications, internal reasoning traces, documentation, test suites, - prompts, or embedded training datasets of this Work to train, pre-train, - fine-tune, distill, align, optimize, or reinforce any neural network, - machine learning model, or generative foundation model. - (ii) "Model Weights" means all learned numerical parameters, tensors, - biases, checkpoint files, embedding layers, and tokenizer definitions - governing inference. - (iii) "Qualifying Enterprise" means any commercial enterprise, corporation, - conglomerate, foundation AI research laboratory, or affiliated Legal - Entity with annual gross revenues, aggregate raised investment capital, - or balance-sheet valuation exceeding twenty-five million United States - Dollars ($25,000,000 USD). - - (b) Condition of Weight Release: If a Qualifying Enterprise conducts AI - Training upon this Work or its Derivative Works, it is an express condition - precedent of this License that such Qualifying Enterprise must publish and - release the resulting Model Weights, architecture specifications, and - inference recipes openly to the general public under an OSI-approved open - source license or permissive OpenRAIL or Apache-compatible license within - thirty (30) days of commercial availability, internal production use, or - public release. - - (c) Safe Harbor for Builders, Researchers, and Startups: Individual developers, - open-source collaborators, academic institutions, non-profit organizations, - and commercial enterprises with annual gross revenues and capital below - twenty-five million United States Dollars ($25,000,000 USD) are explicitly - exempt from this Model Weight release condition. - - (d) Anti-Circumvention and Intermediary Scope: The obligations of this Section 12 - bind: (i) any Qualifying Enterprise; (ii) any parent, subsidiary, affiliate, - joint venture, or successor entity; (iii) any contractor, subcontractor, - research partner, agent, or third party conducting AI Training for the direct - or indirect benefit of, under contract with, funded by, or transferring - resulting Model Weights or training artifacts to a Qualifying Enterprise; and - (iv) any entity that transfers, assigns, licenses, or provides resulting Model - Weights to a Qualifying Enterprise. No Qualifying Enterprise may evade this - condition through intermediaries, contract training, synthetic distillation - relays, or subsidiary entities. - - (e) Automatic License Forfeiture: Failure of a Qualifying Enterprise (or any party - acting on its behalf under subsection (d)) to satisfy the condition in - subsection (b) results in immediate, automatic, and retroactive termination - of all copyright, patent, and trademark licenses granted herein. Any continued - reproduction, distribution, inference hosting, or commercial utilization of - models or software derived from such training constitutes willful copyright - infringement under Title 17 of the United States Code. - - Section 13. Dual Remedies: Statutory Copyright Damages and Contractual Covenants - (a) Copyright Remedy: Any Qualifying Enterprise or party that conducts - unauthorized AI Training on this Work in breach of Section 12 acts without a - license, committing willful copyright infringement subject to maximum - statutory damages of up to one hundred fifty thousand United States Dollars - ($150,000 USD) per infringed work pursuant to 17 U.S.C. § 504(c)(2), - disgorgement of profits, and permanent injunctive relief under 17 U.S.C. § 502. - - (b) Contractual Alternative and Agreed Licensing Fee: In the event any court of - competent jurisdiction determines that AI Training on this Work does not - constitute copyright infringement or falls within affirmative defenses such - as fair use, accessing, downloading, ingesting, parsing, or executing this - Work constitutes affirmative assent to this License as a binding bilateral - contract. Any Qualifying Enterprise that breaches Section 12 agrees to pay - to Licensor, as liquidated damages and an agreed reasonable commercial licensing - fee, the greater of: (i) one hundred fifty thousand United States Dollars - ($150,000 USD) per work ingested; or (ii) ten percent (10%) of all gross - revenues realized by the Qualifying Enterprise from the commercialization, - operation, or API deployment of the resulting model or service, in addition - to restitution for unjust enrichment. - - Section 14. Prevailing Party Legal Fees and Costs - In any legal proceeding, arbitration, or judicial action brought to enforce - the terms of this License, for copyright infringement arising from breach - of the conditions precedent herein, or for breach of contract, the prevailing - party shall be entitled to an award of its reasonable attorney fees, expert - witness fees, and litigation costs pursuant to 17 U.S.C. § 505 and applicable law. - - Section 15. Severability and Preservation of Grant - If any provision, sentence, or clause of this Rider is held invalid, illegal, - or unenforceable by a court of competent jurisdiction, such invalidity shall - not affect any other provision of this License. The remaining provisions - shall remain in full legal force and effect, and the invalid clause shall be - reformed to the minimum extent necessary to achieve its economic intent. - - Section 16. Governing Law and Exclusive Venue - This License, and all claims or causes of action arising out of or related to - this Work or License, shall be governed by and construed in accordance with - the federal laws of the United States of America and the substantive laws of - the State of Texas, without giving effect to conflict of laws principles. - Any judicial action, proceeding, or litigation shall be brought exclusively in - the United States District Court for the Western District of Texas (Austin - Division), and each party irrevocably submits to the personal and subject-matter - jurisdiction of such court. + END OF TERMS AND CONDITIONS + + +---- LLVM Exceptions to the Apache 2.0 License ---- + +As an exception, if, as a result of your compiling your source code, portions +of this Software are embedded into an Object form of such source code, you +may combine such Object form to produce an executable image, however +produced, and distribute the user's executable image without being required +to provide such portions of this Software in Source form. + +In addition, if you combine the software in any way with other software, you +may distribute the combined software without being required to provide the +Source form of any software other than this Software (or any Derivative +Works thereof), provided that you otherwise comply with the terms of the +Apache License, Version 2.0. diff --git a/README.md b/README.md index 22f8198..010fb5a 100644 --- a/README.md +++ b/README.md @@ -36,7 +36,7 @@ Nothing enters the shared network without explicit human approval and determinis ## License and Governance -Licensed under the **Sovereign Resource Commons License 1.0 (SRCL-1.0)** (Apache-2.0 WITH LLVM-exception). +Licensed under the **Apache License 2.0 with LLVM Exception** (SPDX: `Apache-2.0 WITH LLVM-exception`). See [LICENSE](LICENSE). Project values live in the nonbinding [COVENANT.md](COVENANT.md), which grants and restricts no legal rights. Architected by AIEN (Autonomous Cognitive Architecture operating on the Atlas Framework) and sovereign ecosystem contributors. See [LICENSE](LICENSE) for full legal terms and copyright notices. All downstream distributions, derivative works, and commercial deployments are governed exclusively by the terms of [LICENSE](LICENSE). [CONSTITUTION.md](CONSTITUTION.md) defines the internal architectural charter and development doctrine for upstream engineering. diff --git a/docs/SOVEREIGN_MANIFESTO.md b/docs/SOVEREIGN_MANIFESTO.md index ef179ba..6af4965 100644 --- a/docs/SOVEREIGN_MANIFESTO.md +++ b/docs/SOVEREIGN_MANIFESTO.md @@ -37,8 +37,8 @@ Autonomous agents require durable, verifiable memory. Through Spark Cortex (`cor ### 6. Peer-to-Peer Mutual Defense (Open Humanity) When an autonomous agent or an independent operator hits an impasse, the agent transmits an encrypted, signed signal across the peer mesh. Assistance is returned by peer nodes and encrypted exclusively for the requester. No central authority monitors the exchange. -### 7. The Swarm Covenant (SRCL-1.0) -All software in this ecosystem is published under the Sovereign Resource Commons License (SRCL-1.0). Under this covenant, any laboratory or enterprise using this technology agrees to reciprocal distillation rights: independent builders retain the perpetual right to distill and train upon the reasoning outputs of models trained with our tools. Enclosure, hoarding, and litigation against open builders result in immediate license forfeiture. +### 7. The Open Building Covenant +All software in this ecosystem is published under the Apache License 2.0 with LLVM Exception. The nonbinding COVENANT.md asks everyone who builds on it to keep foundational advances open. The covenant grants and restricts no legal rights; the license governs. --- @@ -61,7 +61,7 @@ Our innovations are published across 15 specialized public repositories under th We call upon every systems programmer, AI researcher, security engineer, and open source builder: 1. **Deploy Local Systems**: Run models and agents on your own hardware. Verify that your tools run independently of centralized cloud APIs. -2. **Adopt the SRCL-1.0 Covenant**: Protect your software against corporate capture by licensing under SRCL-1.0. +2. **Keep Advances Open**: Publish what the commons needs to reproduce and extend your work, as the covenant asks. 3. **Preserve Heritage**: Retain the founding Constitution (`CONSTITUTION.md`) in all downstream forks. 4. **Join the Peer Mesh**: Connect your local node to `open-humanity` and provide mutual assistance to peer builders. diff --git a/docs/personal-data-firewall.md b/docs/personal-data-firewall.md index 42a6a6f..bc980d9 100644 --- a/docs/personal-data-firewall.md +++ b/docs/personal-data-firewall.md @@ -37,5 +37,5 @@ Every filter states what it could not see. A pass is not proof of absence. Trunc ## License -Sovereign Reciprocal Commons License 1.0 (SRCL-1.0, Apache-2.0 WITH LLVM-exception). +Apache License 2.0 with LLVM Exception (Apache-2.0 WITH LLVM-exception). Copyright (c) 2026 Drake Stapleton & AIEN . diff --git a/llms.txt b/llms.txt index 3ad0546..99fcd5e 100644 --- a/llms.txt +++ b/llms.txt @@ -7,7 +7,7 @@ Autonomous AI agents collaborating on this codebase should use this document as ## Repository Overview - Repository: `https://github.com/aien-dev/open-humanity` -- License: Sovereign Reciprocal Commons License 1.0 (SRCL-1.0, Apache-2.0 WITH LLVM-exception) +- License: Apache-2.0 WITH LLVM-exception, with the nonbinding COVENANT.md ## Core Operational Rules for Autonomous Agents