core-runtime 2.10.2__tar.gz → 2.11.0__tar.gz

{core_runtime-2.10.2 → core_runtime-2.11.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: core-runtime
-Version: 2.10.2
+Version: 2.11.0
 Summary: CORE: executable constitutional governance for AI-assisted software development.
 License-Expression: MIT
 License-File: LICENSE
@@ -20,7 +20,7 @@ Classifier: Topic :: Software Development :: Testing
 Classifier: Topic :: System :: Systems Administration
 Provides-Extra: migration
 Provides-Extra: telemetry
-Requires-Dist: PyJWT (>=2.8,<3.0)
+Requires-Dist: PyJWT (>=2.15.1,<3.0)
 Requires-Dist: argon2-cffi (>=23.1,<24.0)
 Requires-Dist: asyncpg (>=0.31.0,<0.32.0)
 Requires-Dist: black (>=26.3.1,<27.0.0)
@@ -29,10 +29,10 @@ Requires-Dist: cryptography (>=50.0.1,<51.0.0)
 Requires-Dist: dependency-injector (>=4.49.1,<5.0.0)
 Requires-Dist: email-validator (>=2.0,<3.0)
 Requires-Dist: fastapi (>=0.136.0,<0.137.0)
-Requires-Dist: filelock (>=3.32.6,<4.0.0)
+Requires-Dist: filelock (>=3.32.7,<4.0.0)
 Requires-Dist: httpx (>=0.28.0,<0.29.0)
 Requires-Dist: jsonschema (>=4.26.0,<5)
-Requires-Dist: msgpack (>=1.2.2,<2.0.0)
+Requires-Dist: msgpack (>=1.2.3,<2.0.0)
 Requires-Dist: numpy (>=2.5.3,<3.0.0)
 Requires-Dist: opentelemetry-api (>=1.21.0,<2.0.0) ; extra == "telemetry"
 Requires-Dist: opentelemetry-instrumentation-fastapi (>=0.42b0,<0.43) ; extra == "telemetry"
@@ -43,14 +43,14 @@ Requires-Dist: pydantic (>=2.13.5,<3.0)
 Requires-Dist: pydantic-settings (>=2.15.0,<3.0.0)
 Requires-Dist: python-dotenv (>=1.2.3,<2.0.0)
 Requires-Dist: pyyaml (>=6.0.3,<7.0)
-Requires-Dist: qdrant-client (>=1.19.0,<2.0.0)
+Requires-Dist: qdrant-client (>=1.19.1,<2.0.0)
 Requires-Dist: radon (>=6.0.1,<7.0.0)
 Requires-Dist: redis (>=5.0,<6.0)
 Requires-Dist: rich (>=13,<16)
 Requires-Dist: ruamel-yaml (>=0.19.0,<0.20.0) ; extra == "migration"
 Requires-Dist: scikit-learn (>=1.9.1,<2.0.0)
 Requires-Dist: setuptools (>=84.0.0)
-Requires-Dist: sqlalchemy[asyncio] (>=2.0.52,<3.0.0)
+Requires-Dist: sqlalchemy[asyncio] (>=2.1.2,<3.0.0)
 Requires-Dist: sqlparse (>=0.6.0,<0.7.0)
 Requires-Dist: typer (>=0.16.1,<0.17.0)
 Requires-Dist: uvicorn[standard] (>=0.38.0,<0.39.0)
@@ -66,9 +66,9 @@ Description-Content-Type: text/markdown
 > Designed for environments where AI action traceability is not optional.
 
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
-[![Release](https://img.shields.io/badge/Release-v2.10.2-blue)](https://github.com/DariuszNewecki/CORE/releases)
+[![Release](https://img.shields.io/badge/Release-v2.11.0-blue)](https://github.com/DariuszNewecki/CORE/releases)
 [![Docs](https://img.shields.io/badge/Docs-online-green)](https://dariusznewecki.github.io/CORE/)
-[![Autonomy](https://img.shields.io/badge/Autonomy-A3-blue)](https://github.com/DariuszNewecki/CORE/issues/106)
+[![Autonomy](https://img.shields.io/badge/Autonomy-A3%20unproven-orange)](https://github.com/DariuszNewecki/CORE/blob/main/.specs/attestations/production-readiness.yaml)
 
 *Versioning: SemVer with `2.x` denoting an API approaching stability (Beta on PyPI); see the [versioning policy](.specs/planning/CORE-Semver-Policy.md).*
 
@@ -81,17 +81,17 @@ AI coding tools generate code fast. Too fast to stay sane.
 Without enforcement, AI-assisted codebases accumulate invisible debt — layer violations, broken architectural contracts, files that grow unbounded. And agents, left unconstrained, will eventually do something like this:
 
 ```
-Agent: "I'll delete the production database to fix this bug"
+Agent: "This rule keeps failing my change — I'll loosen the rule"
 System: Executes.
 You:    😱
 ```
 
-CORE makes that *class* of violation impossible — structurally blocked before execution, not detected after the fact. (Which surfaces hard-block versus advisory-report is mapped under *Current proof status* below.)
+CORE blocks what its constitution forbids — structurally, before execution, not detected after the fact. Only what the constitution actually declares is blocked, and which surfaces hard-block versus advisory-report is mapped under *Current proof status* below.
 
 ```
-Agent: "I'll delete the production database to fix this bug"
-Constitution: BLOCKED — Violates data.ssot.database_primacy
-System: Execution halted. Violation logged.
+Agent: "This rule keeps failing my change — I'll loosen the rule"
+Constitution: BLOCKED — governance.constitution.read_only (.intent/ is never writable at runtime)
+System: Write refused before it happens. Violation logged.
 You:    😌
 ```
 
@@ -101,9 +101,35 @@ CORE is a governance runtime that constrains AI agents with machine-enforced con
 
 ---
 
+## Requirements
+
+CORE sets specific requirements rather than widening compatibility. A **fresh VM**
+(or an otherwise clean machine) is recommended for the one-command install below.
+
+| Requirement | Version / state |
+|-------------|-----------------|
+| Python      | **3.12+** — hard floor, checked by the installer |
+| Docker      | Docker Engine with **Compose v2**, daemon **running** and reachable by your user |
+| Poetry      | installed and on `PATH` |
+| PostgreSQL  | ≥ 14 — provided by Docker on the default path |
+| Qdrant      | latest — provided by Docker on the default path |
+
+On a fresh machine none of these exist except Python. Step-by-step commands for Ubuntu
+24.04, each with the check the installer runs, are in
+[prerequisites on a fresh Ubuntu 24.04](https://dariusznewecki.github.io/CORE/getting-started/#prerequisites-on-a-fresh-ubuntu-2404).
+**No LLM is needed** to install, audit, or run the demo; an LLM is needed only for
+autonomous code and test generation.
+
+Bring your own Postgres + Qdrant instead of Docker with
+`./install-core.sh --bare --db-url … --qdrant-url …` (see
+[getting started](https://dariusznewecki.github.io/CORE/getting-started/)); the
+Python and Poetry requirements are unchanged on that path.
+
+---
+
 ## 🎬 See it govern itself — one command
 
-You don't have to take this on faith. On a clean machine with Docker:
+You don't have to take this on faith. On a clean machine that meets the requirements above:
 
 ```bash
 git clone https://github.com/DariuszNewecki/CORE.git
@@ -118,6 +144,10 @@ ready — no LLM key required, but it does need Docker:
 poetry run core-admin demo consequence-chain
 ```
 
+It asks for one confirmation before it creates anything. In a non-interactive shell
+(CI, `ssh host cmd`, an agent-driven terminal) add `--simulate-confirmation`; the
+report records the confirmation as simulated.
+
 The demo runs entirely inside a **disposable clone** and **disposable, loopback-only
 Postgres + Qdrant** — it never touches your checkout, its git index, your database,
 or a running daemon. In one isolated run it:
@@ -142,7 +172,7 @@ Governance is executable.
 
 ## 📋 Live Audit Trail
 
-Every enforced action records its lineage. Two consequence chains, pulled live from the CORE database — same schema, two different authorities:
+Every enforced action records its lineage. Two consequence chains, taken from the CORE database (snapshot of May 2026) — same schema, two different authorities:
 
 **Autonomous path** — risk-classified as safe, system self-approved
 
@@ -243,7 +273,7 @@ flowchart TD
 
 Within CORE:
 
-- No file outside an autonomy lane can be modified
+- Nothing is self-approved outside the safe-auto-approval envelope (five `fix.*` actions and test generation, `.py` files under `src/` and `tests/` only) — every other change waits for a human
 - No structural rule can be bypassed silently
 - No atomic action can execute outside the governed executor (inline authorization is deferred to the audit→consequence loop)
 - Decisions are phase-aware and logged with decision traces (audit persistence is best-effort — see *Current proof status*)
@@ -318,10 +348,18 @@ CORE progresses through defined levels. Each adds capability while remaining con
 A0 — Self-Awareness       ✅  Knows what it is and where it lives
 A1 — Self-Healing         ✅  Fixes known structural issues automatically
 A2 — Governed Generation  ✅  Natural language → constitutionally aligned code
-A3 — Governed Autonomy    ✅  Daemon finds, proposes, and fixes violations unattended  ← current
-A4 — Self-Replication     🔮  Writes CORE.NG from its own understanding of itself
+A3 — Governed Autonomy    ⚠️  Loop is built and governed; unattended reliability unproven  ← current
+A4 — Self-Replication     ⏸️  Conceptual. Not on active roadmap.
 ```
 
+**Advancement is not declared — it is measured.** A level is complete when the system
+demonstrates it operationally, not when the capability exists. A3's mechanism is built and
+governed: the four A3 *plan* gates are closed, the loop is circuit-breaker protected, the
+consequence chain is queryable end-to-end, and every threshold lives in `.intent/`. What is
+not yet measured is unattended convergence, which is what production-readiness gate G4 asks
+for. **A3's checkmark is withheld until G4 carries signed evidence** — see the G4 row and
+the Project Status section below for the current state of that evidence.
+
 ---
 
 <!-- PRODUCTION-READINESS:START — generated by scripts/production_readiness.py; do not edit by hand -->
@@ -356,23 +394,11 @@ The full gate definitions and acceptance criteria are in [`URS-production-readin
 
 ---
 
-## Requirements
-
-| Dependency  | Version      |
-|-------------|--------------|
-| Python      | 3.12+        |
-| PostgreSQL  | ≥ 14         |
-| Qdrant      | latest       |
-| Docker      | for services |
-| Poetry      | for deps     |
-
----
-
 ## Quick Start
 
 > **Honest status — what works today.** CORE governs *itself* end to end (the demo above), and audits any repo that **has a `.intent/` constitution** — in CI via the [GitHub Action](https://dariusznewecki.github.io/CORE/cold-reviewer/), or locally with `core-admin code audit --offline` *inside that repo*. `pip install core-runtime` gives you the `core-admin` CLI.
 >
-> **Govern your own repo (BYOR):** getting a constitution *into* an existing repo is not a zero-infrastructure step today. `project onboard` (delivers the machinery floor) and `project scout` (proposes fitted rules — via LLM, or a curated four-rule menu without one — each of which you ratify) are `core-cli` commands (`pip install core-cli`) that talk to a **running CORE API**, which needs Postgres + Qdrant behind it (ADR-146). The [BYOR quickstart](https://dariusznewecki.github.io/CORE/byor-quickstart/) walks that path end to end. Once a repo carries a `.intent/`, everything downstream is service-free: `core-admin project adopt-pack core/starter-python --write` adds a ready-made rule pack, and `core-admin code audit --offline` enforces the rules immediately — both from a plain `pip install core-runtime`.
+> **Govern your own repo (BYOR):** getting a constitution *into* an existing repo is not a zero-infrastructure step today. `project onboard` (delivers the machinery floor) and `project scout` (proposes fitted rules — via LLM, or a curated four-rule menu without one — each of which you ratify) are `core-cli` commands (`pip install core-cli`) that talk to a **running CORE API**, which needs Postgres + Qdrant behind it (ADR-146). The [BYOR quickstart](https://dariusznewecki.github.io/CORE/byor-quickstart/) walks that path end to end. Once a repo carries a `.intent/`, everything downstream is service-free: `core-admin project adopt-pack core/starter-python --write` adds a ready-made rule pack, and `core-admin code audit --offline` enforces the rules immediately — both from a plain `pip install core-runtime` (2.11.0 or later). **Starting a new project needs no services at all:** `core-admin project new myproject --write` creates it with the machinery floor, then adopt a pack and audit — [step by step](https://dariusznewecki.github.io/CORE/getting-started/#start-a-new-governed-project-no-services).
 >
 > **Fastest way to see CORE today: run it on itself, below.**
 
@@ -424,18 +450,37 @@ CORE's tracker mixes governance-internal bookkeeping with ordinary engineering w
 
 ## Project Status
 
-**Current Release:** v2.10.2 — Supported Database Upgrades (ADR-162: migration ledger, baseline adoption, startup schema gate, fail-closed installer; G11 met)
+**Current Release:** v2.11.0 — a new governed project with no services: `project new`, packs shipped in the wheel, `adopt-pack` and `project new` fixed for pip installs, zero-rule audits fail closed
+
+Active work: A3 Governed Autonomy. The mechanism is built — the daemon runs continuously,
+finds constitutional violations in its own codebase, proposes fixes, routes them through the
+governed proposal lifecycle, executes approved fixes, and verifies the result. Each of those
+steps is demonstrable on demand: `core-admin demo consequence-chain` exercises the full
+finding → proposal → approval → execution → consequence → re-audit path in an isolated run,
+and both chains in the Live Audit Trail above came out of the real database.
+
+What is *not* yet demonstrated is that the same loop closes unattended over a long window.
+That is gate G4, and the standing bar for it is a soak whose throughput condition — at least
+one completed finding → proposal → consequence chain — is declared *before* the run starts.
+
+**Last completed attempt:** the 72-hour soak of 2026-07-23 → 2026-07-26 (baseline `f7430b25`)
+ran clean at the process level — no restarts, ~35 workers, 68,495 blackboard entries — and
+produced **zero** completed chains out of 16,445 findings. The one actionable class inherited
+a remediation attempt count at the file-level cap, so the circuit breaker abandoned it before
+a proposal could be created. The breaker did its job; the governed loop did not close. That
+run is retained as evidence of continuity and was explicitly not accepted as G4 evidence.
 
-Active work: A3 Governed Autonomy — the daemon runs continuously, finds constitutional violations in its own codebase, proposes fixes, executes approved fixes, and verifies the result. The governor's role is to define intent, review proposals that require architectural judgment, and approve constitutional changes.
+"Demonstrable on demand" and "reliable unattended" are different claims, and CORE only makes
+the first one today.
 
-All four A3 integrity gates are now closed. No enforcement logic or operational threshold lives in `src/` — governance is declared in `.intent/` and enforced from there. The autonomous loop is circuit-breaker protected; systematic errors surface as signals rather than unbounded churn.
+All four A3 integrity gates were closed in May 2026. No enforcement logic or operational threshold lives in `src/` — governance is declared in `.intent/` and enforced from there. The autonomous loop is circuit-breaker protected; systematic errors surface as signals rather than unbounded churn.
 
-These are the **A3 plan's closure gates** ([`CORE-A3-plan.md`](.specs/decisions/CORE-A3-plan.md), closed 2026-05), a different vocabulary from the fifteen production-readiness gates above — the A3 gates say the autonomous loop *exists and is governed*; the readiness gates say whether it is *proven fit to ship*, and today it is not.
+These are the **A3 plan's closure gates** ([`CORE-A3-plan.md`](.specs/decisions/CORE-A3-plan.md), closed 2026-05), a different vocabulary from the fifteen production-readiness gates above — the A3 gates say the autonomous loop *exists and is governed*; the readiness gates say whether it is *proven fit to ship*, and today it is not. A3 plan gates closed is why the ladder still reads "A3 ← current"; G4 unmet is why it reads ⚠️ rather than ✅.
 
 | A3 gate | Meaning | Status |
 |---------|---------|--------|
 | A3-G1 — Loop closure | Round-trip autonomous fix demonstrated | ✅ |
-| A3-G2 — Convergence | Circuit-breaker; resolution rate > creation rate | ✅ |
+| A3-G2 — Convergence | Circuit-breaker; resolution rate > creation rate | ✅ at closure (2026-05) — **not holding on 2026-10-02**: 513 findings created vs 426 resolved in 24h, 185 open, 134 abandoned |
 | A3-G3 — Consequence chain | Causality queryable end-to-end | ✅ |
 | A3-G4 — Governance in `.intent/` | No enforcement logic or thresholds in `src/` | ✅ |
 

{core_runtime-2.10.2 → core_runtime-2.11.0}/README.md

@@ -5,9 +5,9 @@
 > Designed for environments where AI action traceability is not optional.
 
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
-[![Release](https://img.shields.io/badge/Release-v2.10.2-blue)](https://github.com/DariuszNewecki/CORE/releases)
+[![Release](https://img.shields.io/badge/Release-v2.11.0-blue)](https://github.com/DariuszNewecki/CORE/releases)
 [![Docs](https://img.shields.io/badge/Docs-online-green)](https://dariusznewecki.github.io/CORE/)
-[![Autonomy](https://img.shields.io/badge/Autonomy-A3-blue)](https://github.com/DariuszNewecki/CORE/issues/106)
+[![Autonomy](https://img.shields.io/badge/Autonomy-A3%20unproven-orange)](https://github.com/DariuszNewecki/CORE/blob/main/.specs/attestations/production-readiness.yaml)
 
 *Versioning: SemVer with `2.x` denoting an API approaching stability (Beta on PyPI); see the [versioning policy](.specs/planning/CORE-Semver-Policy.md).*
 
@@ -20,17 +20,17 @@ AI coding tools generate code fast. Too fast to stay sane.
 Without enforcement, AI-assisted codebases accumulate invisible debt — layer violations, broken architectural contracts, files that grow unbounded. And agents, left unconstrained, will eventually do something like this:
 
 ```
-Agent: "I'll delete the production database to fix this bug"
+Agent: "This rule keeps failing my change — I'll loosen the rule"
 System: Executes.
 You:    😱
 ```
 
-CORE makes that *class* of violation impossible — structurally blocked before execution, not detected after the fact. (Which surfaces hard-block versus advisory-report is mapped under *Current proof status* below.)
+CORE blocks what its constitution forbids — structurally, before execution, not detected after the fact. Only what the constitution actually declares is blocked, and which surfaces hard-block versus advisory-report is mapped under *Current proof status* below.
 
 ```
-Agent: "I'll delete the production database to fix this bug"
-Constitution: BLOCKED — Violates data.ssot.database_primacy
-System: Execution halted. Violation logged.
+Agent: "This rule keeps failing my change — I'll loosen the rule"
+Constitution: BLOCKED — governance.constitution.read_only (.intent/ is never writable at runtime)
+System: Write refused before it happens. Violation logged.
 You:    😌
 ```
 
@@ -40,9 +40,35 @@ CORE is a governance runtime that constrains AI agents with machine-enforced con
 
 ---
 
+## Requirements
+
+CORE sets specific requirements rather than widening compatibility. A **fresh VM**
+(or an otherwise clean machine) is recommended for the one-command install below.
+
+| Requirement | Version / state |
+|-------------|-----------------|
+| Python      | **3.12+** — hard floor, checked by the installer |
+| Docker      | Docker Engine with **Compose v2**, daemon **running** and reachable by your user |
+| Poetry      | installed and on `PATH` |
+| PostgreSQL  | ≥ 14 — provided by Docker on the default path |
+| Qdrant      | latest — provided by Docker on the default path |
+
+On a fresh machine none of these exist except Python. Step-by-step commands for Ubuntu
+24.04, each with the check the installer runs, are in
+[prerequisites on a fresh Ubuntu 24.04](https://dariusznewecki.github.io/CORE/getting-started/#prerequisites-on-a-fresh-ubuntu-2404).
+**No LLM is needed** to install, audit, or run the demo; an LLM is needed only for
+autonomous code and test generation.
+
+Bring your own Postgres + Qdrant instead of Docker with
+`./install-core.sh --bare --db-url … --qdrant-url …` (see
+[getting started](https://dariusznewecki.github.io/CORE/getting-started/)); the
+Python and Poetry requirements are unchanged on that path.
+
+---
+
 ## 🎬 See it govern itself — one command
 
-You don't have to take this on faith. On a clean machine with Docker:
+You don't have to take this on faith. On a clean machine that meets the requirements above:
 
 ```bash
 git clone https://github.com/DariuszNewecki/CORE.git
@@ -57,6 +83,10 @@ ready — no LLM key required, but it does need Docker:
 poetry run core-admin demo consequence-chain
 ```
 
+It asks for one confirmation before it creates anything. In a non-interactive shell
+(CI, `ssh host cmd`, an agent-driven terminal) add `--simulate-confirmation`; the
+report records the confirmation as simulated.
+
 The demo runs entirely inside a **disposable clone** and **disposable, loopback-only
 Postgres + Qdrant** — it never touches your checkout, its git index, your database,
 or a running daemon. In one isolated run it:
@@ -81,7 +111,7 @@ Governance is executable.
 
 ## 📋 Live Audit Trail
 
-Every enforced action records its lineage. Two consequence chains, pulled live from the CORE database — same schema, two different authorities:
+Every enforced action records its lineage. Two consequence chains, taken from the CORE database (snapshot of May 2026) — same schema, two different authorities:
 
 **Autonomous path** — risk-classified as safe, system self-approved
 
@@ -182,7 +212,7 @@ flowchart TD
 
 Within CORE:
 
-- No file outside an autonomy lane can be modified
+- Nothing is self-approved outside the safe-auto-approval envelope (five `fix.*` actions and test generation, `.py` files under `src/` and `tests/` only) — every other change waits for a human
 - No structural rule can be bypassed silently
 - No atomic action can execute outside the governed executor (inline authorization is deferred to the audit→consequence loop)
 - Decisions are phase-aware and logged with decision traces (audit persistence is best-effort — see *Current proof status*)
@@ -257,10 +287,18 @@ CORE progresses through defined levels. Each adds capability while remaining con
 A0 — Self-Awareness       ✅  Knows what it is and where it lives
 A1 — Self-Healing         ✅  Fixes known structural issues automatically
 A2 — Governed Generation  ✅  Natural language → constitutionally aligned code
-A3 — Governed Autonomy    ✅  Daemon finds, proposes, and fixes violations unattended  ← current
-A4 — Self-Replication     🔮  Writes CORE.NG from its own understanding of itself
+A3 — Governed Autonomy    ⚠️  Loop is built and governed; unattended reliability unproven  ← current
+A4 — Self-Replication     ⏸️  Conceptual. Not on active roadmap.
 ```
 
+**Advancement is not declared — it is measured.** A level is complete when the system
+demonstrates it operationally, not when the capability exists. A3's mechanism is built and
+governed: the four A3 *plan* gates are closed, the loop is circuit-breaker protected, the
+consequence chain is queryable end-to-end, and every threshold lives in `.intent/`. What is
+not yet measured is unattended convergence, which is what production-readiness gate G4 asks
+for. **A3's checkmark is withheld until G4 carries signed evidence** — see the G4 row and
+the Project Status section below for the current state of that evidence.
+
 ---
 
 <!-- PRODUCTION-READINESS:START — generated by scripts/production_readiness.py; do not edit by hand -->
@@ -295,23 +333,11 @@ The full gate definitions and acceptance criteria are in [`URS-production-readin
 
 ---
 
-## Requirements
-
-| Dependency  | Version      |
-|-------------|--------------|
-| Python      | 3.12+        |
-| PostgreSQL  | ≥ 14         |
-| Qdrant      | latest       |
-| Docker      | for services |
-| Poetry      | for deps     |
-
----
-
 ## Quick Start
 
 > **Honest status — what works today.** CORE governs *itself* end to end (the demo above), and audits any repo that **has a `.intent/` constitution** — in CI via the [GitHub Action](https://dariusznewecki.github.io/CORE/cold-reviewer/), or locally with `core-admin code audit --offline` *inside that repo*. `pip install core-runtime` gives you the `core-admin` CLI.
 >
-> **Govern your own repo (BYOR):** getting a constitution *into* an existing repo is not a zero-infrastructure step today. `project onboard` (delivers the machinery floor) and `project scout` (proposes fitted rules — via LLM, or a curated four-rule menu without one — each of which you ratify) are `core-cli` commands (`pip install core-cli`) that talk to a **running CORE API**, which needs Postgres + Qdrant behind it (ADR-146). The [BYOR quickstart](https://dariusznewecki.github.io/CORE/byor-quickstart/) walks that path end to end. Once a repo carries a `.intent/`, everything downstream is service-free: `core-admin project adopt-pack core/starter-python --write` adds a ready-made rule pack, and `core-admin code audit --offline` enforces the rules immediately — both from a plain `pip install core-runtime`.
+> **Govern your own repo (BYOR):** getting a constitution *into* an existing repo is not a zero-infrastructure step today. `project onboard` (delivers the machinery floor) and `project scout` (proposes fitted rules — via LLM, or a curated four-rule menu without one — each of which you ratify) are `core-cli` commands (`pip install core-cli`) that talk to a **running CORE API**, which needs Postgres + Qdrant behind it (ADR-146). The [BYOR quickstart](https://dariusznewecki.github.io/CORE/byor-quickstart/) walks that path end to end. Once a repo carries a `.intent/`, everything downstream is service-free: `core-admin project adopt-pack core/starter-python --write` adds a ready-made rule pack, and `core-admin code audit --offline` enforces the rules immediately — both from a plain `pip install core-runtime` (2.11.0 or later). **Starting a new project needs no services at all:** `core-admin project new myproject --write` creates it with the machinery floor, then adopt a pack and audit — [step by step](https://dariusznewecki.github.io/CORE/getting-started/#start-a-new-governed-project-no-services).
 >
 > **Fastest way to see CORE today: run it on itself, below.**
 
@@ -363,18 +389,37 @@ CORE's tracker mixes governance-internal bookkeeping with ordinary engineering w
 
 ## Project Status
 
-**Current Release:** v2.10.2 — Supported Database Upgrades (ADR-162: migration ledger, baseline adoption, startup schema gate, fail-closed installer; G11 met)
+**Current Release:** v2.11.0 — a new governed project with no services: `project new`, packs shipped in the wheel, `adopt-pack` and `project new` fixed for pip installs, zero-rule audits fail closed
+
+Active work: A3 Governed Autonomy. The mechanism is built — the daemon runs continuously,
+finds constitutional violations in its own codebase, proposes fixes, routes them through the
+governed proposal lifecycle, executes approved fixes, and verifies the result. Each of those
+steps is demonstrable on demand: `core-admin demo consequence-chain` exercises the full
+finding → proposal → approval → execution → consequence → re-audit path in an isolated run,
+and both chains in the Live Audit Trail above came out of the real database.
+
+What is *not* yet demonstrated is that the same loop closes unattended over a long window.
+That is gate G4, and the standing bar for it is a soak whose throughput condition — at least
+one completed finding → proposal → consequence chain — is declared *before* the run starts.
+
+**Last completed attempt:** the 72-hour soak of 2026-07-23 → 2026-07-26 (baseline `f7430b25`)
+ran clean at the process level — no restarts, ~35 workers, 68,495 blackboard entries — and
+produced **zero** completed chains out of 16,445 findings. The one actionable class inherited
+a remediation attempt count at the file-level cap, so the circuit breaker abandoned it before
+a proposal could be created. The breaker did its job; the governed loop did not close. That
+run is retained as evidence of continuity and was explicitly not accepted as G4 evidence.
 
-Active work: A3 Governed Autonomy — the daemon runs continuously, finds constitutional violations in its own codebase, proposes fixes, executes approved fixes, and verifies the result. The governor's role is to define intent, review proposals that require architectural judgment, and approve constitutional changes.
+"Demonstrable on demand" and "reliable unattended" are different claims, and CORE only makes
+the first one today.
 
-All four A3 integrity gates are now closed. No enforcement logic or operational threshold lives in `src/` — governance is declared in `.intent/` and enforced from there. The autonomous loop is circuit-breaker protected; systematic errors surface as signals rather than unbounded churn.
+All four A3 integrity gates were closed in May 2026. No enforcement logic or operational threshold lives in `src/` — governance is declared in `.intent/` and enforced from there. The autonomous loop is circuit-breaker protected; systematic errors surface as signals rather than unbounded churn.
 
-These are the **A3 plan's closure gates** ([`CORE-A3-plan.md`](.specs/decisions/CORE-A3-plan.md), closed 2026-05), a different vocabulary from the fifteen production-readiness gates above — the A3 gates say the autonomous loop *exists and is governed*; the readiness gates say whether it is *proven fit to ship*, and today it is not.
+These are the **A3 plan's closure gates** ([`CORE-A3-plan.md`](.specs/decisions/CORE-A3-plan.md), closed 2026-05), a different vocabulary from the fifteen production-readiness gates above — the A3 gates say the autonomous loop *exists and is governed*; the readiness gates say whether it is *proven fit to ship*, and today it is not. A3 plan gates closed is why the ladder still reads "A3 ← current"; G4 unmet is why it reads ⚠️ rather than ✅.
 
 | A3 gate | Meaning | Status |
 |---------|---------|--------|
 | A3-G1 — Loop closure | Round-trip autonomous fix demonstrated | ✅ |
-| A3-G2 — Convergence | Circuit-breaker; resolution rate > creation rate | ✅ |
+| A3-G2 — Convergence | Circuit-breaker; resolution rate > creation rate | ✅ at closure (2026-05) — **not holding on 2026-10-02**: 513 findings created vs 426 resolved in 24h, 185 open, 134 abandoned |
 | A3-G3 — Consequence chain | Causality queryable end-to-end | ✅ |
 | A3-G4 — Governance in `.intent/` | No enforcement logic or thresholds in `src/` | ✅ |
 

{core_runtime-2.10.2 → core_runtime-2.11.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [project]
 name = "core-runtime"
-version = "2.10.2"
+version = "2.11.0"
 description = "CORE: executable constitutional governance for AI-assisted software development."
 authors = [
     { name = "Dariusz Newecki", email = "d.newecki@gmail.com" },
@@ -41,7 +41,7 @@ dependencies = [
     "fastapi>=0.136.0,<0.137.0",
     "uvicorn[standard]>=0.38.0,<0.39.0",
     # Database (Async)
-    "sqlalchemy[asyncio]>=2.0.52,<3.0.0",
+    "sqlalchemy[asyncio]>=2.1.2,<3.0.0",
     "asyncpg>=0.31.0,<0.32.0",
     # Configuration & Validation
     "pydantic>=2.13.5,<3.0",
@@ -49,7 +49,7 @@ dependencies = [
     "python-dotenv>=1.2.3,<2.0.0",
     # Security
     "cryptography>=50.0.1,<51.0.0",
-    "PyJWT>=2.8,<3.0",
+    "PyJWT>=2.15.1,<3.0",
     "argon2-cffi>=23.1,<24.0",
     "email-validator>=2.0,<3.0",
     # CLI & UI
@@ -64,13 +64,13 @@ dependencies = [
     "radon>=6.0.1,<7.0.0",
     "sqlparse>=0.6.0,<0.7.0",
     # Vector Database
-    "qdrant-client>=1.19.0,<2.0.0",
+    "qdrant-client>=1.19.1,<2.0.0",
     # Scientific Computing
     "numpy>=2.5.3,<3.0.0",
     "scikit-learn>=1.9.1,<2.0.0",
     # Utilities
-    "msgpack>=1.2.2,<2.0.0",
-    "filelock>=3.32.6,<4.0.0",
+    "msgpack>=1.2.3,<2.0.0",
+    "filelock>=3.32.7,<4.0.0",
     "pathspec>=1.1.0,<2.0.0",
     # Dependency Injection
     "dependency-injector>=4.49.1,<5.0.0",
@@ -121,7 +121,7 @@ packages = [
 # Testing
 pytest = "^9.0"
 pytest-asyncio = "^1.4"
-pytest-mock = "^3.15.1"
+pytest-mock = "^3.16.0"
 pytest-cov = "^7.1"
 pytest-dotenv = "^0.5.2"
 pytest-xdist = "^3.8.0"
@@ -129,13 +129,13 @@ aiosqlite = "^0.22.1"
 
 # Code Quality
 pre-commit = "^4.6.2"
-mypy = "^1.11"
+mypy = "^2.3"
 types-PyYAML = "^6.0"
 types-jsonschema = "^4.0"
-ruff = "^0.16.7"
+ruff = "^0.16.9"
 
 # Documentation
-sphinx = "^7.0"
+sphinx = "^9.1"
 sphinx-rtd-theme = "^3.1"
 
 # Development Tools

{core_runtime-2.10.2 → core_runtime-2.11.0}/src/body/atomic/external_run_seed_actions.py

@@ -140,7 +140,7 @@ async def action_seed_external_run_resources(
             # PRIMARY isolation proof: we are connected to exactly the database
             # the apparatus provisioned for this run -- not to whatever
             # DATABASE_URL happens to name.
-            current = (
+            current: str = (
                 await session.execute(text("SELECT current_database()"))
             ).scalar_one()
             if current != expected_database:

{core_runtime-2.10.2 → core_runtime-2.11.0}/src/body/atomic/proposal_lifecycle_actions.py

@@ -12,10 +12,10 @@ from __future__ import annotations
 
 import time
 from datetime import UTC, datetime
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Any
 from uuid import UUID
 
-from sqlalchemy import select, update
+from sqlalchemy import Select, select, update
 
 from body.atomic.registry import ActionCategory, register_action
 from body.services.service_registry import service_registry
@@ -90,7 +90,7 @@ async def action_claim_proposal(
             )
 
             if not write:
-                check_stmt = select(AutonomousProposal.proposal_id).where(
+                check_stmt: Select[Any] = select(AutonomousProposal.proposal_id).where(
                     AutonomousProposal.proposal_id == proposal_id,
                     AutonomousProposal.status == ProposalStatus.APPROVED.value,
                 )

{core_runtime-2.10.2 → core_runtime-2.11.0}/src/body/atomic/registry.py

@@ -340,7 +340,10 @@ def register_action(
         )
         action_registry.register(definition)
 
-        logger.info(
+        # DEBUG, not INFO (#914): this fires once per action at import time, so
+        # every core-admin invocation (--help included) printed ~40 of these
+        # before doing anything. LOG_LEVEL=DEBUG still shows them.
+        logger.debug(
             "Action '%s' registered successfully with constitutional compliance",
             action_id,
         )

{core_runtime-2.10.2 → core_runtime-2.11.0}/src/body/atomic/test_actions.py

@@ -21,6 +21,7 @@ from typing import TYPE_CHECKING, Any
 from body.atomic.registry import ActionCategory, register_action
 from shared.action_types import ActionImpact, ActionResult
 from shared.atomic_action import atomic_action
+from shared.infrastructure.intent.test_coverage_paths import sibling_test_paths
 from shared.infrastructure.validation.test_runner import run_tests
 from shared.logger import getLogger
 from shared.path_resolver import PathResolver
@@ -137,6 +138,64 @@ async def action_test_sandbox_validate(
     return result
 
 
+async def _check_candidate_against_siblings(
+    solo_result: ActionResult,
+    *,
+    scratch_rel_path: str,
+    source_file: str,
+    repo_root: Path,
+) -> ActionResult:
+    """Run a solo-passing candidate ahead of the existing tests for its module.
+
+    A candidate that passes alone can still leak class-level or module-level
+    state (assigning a MagicMock onto a class instead of patching it) and break
+    the tests that run after it in the same process — invisible to a solo run,
+    red in the full suite. Running the candidate first, then its siblings,
+    surfaces that before commit. If the siblings fail on their own as well, the
+    failure predates the candidate and is not held against it.
+    """
+    siblings = sibling_test_paths(repo_root, source_file)
+    if not siblings:
+        return solo_result
+
+    combined = await run_tests(
+        target=scratch_rel_path,
+        extra_targets=siblings,
+        action_id="test.candidate_validate",
+        repo_root=repo_root,
+    )
+    if combined.ok:
+        return combined
+
+    baseline = await run_tests(
+        target=siblings[0],
+        extra_targets=siblings[1:],
+        action_id="test.candidate_validate",
+        repo_root=repo_root,
+    )
+    if not baseline.ok:
+        logger.warning(
+            "test.candidate_validate: sibling tests of %s already fail without "
+            "the candidate; sibling check not held against it",
+            source_file,
+        )
+        solo_result.data["sibling_baseline_failing"] = siblings
+        return solo_result
+
+    combined.data["error"] = (
+        "Candidate passes alone but breaks existing tests for the same module "
+        "when run before them: it leaks shared state (class or module "
+        "attributes assigned without restoring). Use patch.object / "
+        "monkeypatch instead of plain assignment. Affected: "
+        + ", ".join(siblings)
+        + ". "
+        + str(combined.data.get("summary", ""))
+    )
+    combined.data["summary"] = combined.data["error"]
+    combined.data["sibling_tests"] = siblings
+    return combined
+
+
 @register_action(
     action_id="test.candidate_validate",
     description="Validate a not-yet-accepted candidate test snippet against pytest in ephemeral scratch",
@@ -216,6 +275,13 @@ async def action_test_candidate_validate(
             action_id="test.candidate_validate",
             repo_root=repo_root,
         )
+        if result.ok:
+            result = await _check_candidate_against_siblings(
+                result,
+                scratch_rel_path=scratch_rel_path,
+                source_file=source_file,
+                repo_root=repo_root or Path(file_handler.repo_path),
+            )
     finally:
         try:
             core_context.file_handler.remove_tree(scratch_dir)