agentic-systems 2.1.1__tar.gz → 2.1.2__tar.gz

{agentic_systems-2.1.1 → agentic_systems-2.1.2}/CHANGELOG.md

@@ -1,5 +1,40 @@
 # Changelog
 
+## 2.1.2 - 2026-10-22
+
+The 2.1 grammar and execution contracts remain stable. This delivery hardens
+publication and aligns versioned notebooks, Studio and validation bundles.
+
+### Changed
+
+- The semantic evaluator exposes identical, flat criterion arguments through
+  its public Tool schema and OpenAI Agents callable, preserving the strict
+  evidence-backed rubric across Frameworks.
+- Public release acceptance is scoped to OpenAI, Bedrock API key and Ollama
+  across four Frameworks, with deterministic Python controls. vLLM/Colab,
+  AWS IAM and ADA IAM require subsequent exact-wheel validation. ADA bundles
+  are separate from the mandatory public release assets.
+- Release inference uses a persistent allocation capped at USD 10 and 1000
+  transport attempts, including pilots, evaluators and retries. Reservations
+  are conservative bounds, not billed costs.
+
+- OpenAI Agents preserves completed function-tool evidence when a subsequent
+  SDK operation fails, in both synchronous and asynchronous execution. Recording
+  is execution-local and snapshots outputs without changing the public API.
+
+- Pydantic is constrained to the supported v2 major (`>=2.7.0,<3`).
+- Publication uses typed, hash-addressed manifests and sealed candidate assets;
+  a matching index is a no-upload repeat, while partial or conflicting files
+  fail closed. No published version, tag or asset is overwritten.
+- GitHub finalization verifies uploaded bytes before closing a draft and supports
+  an exact no-op repeat. OIDC jobs and environments have scoped permissions.
+- Quality CI tests both Python 3.10 and 3.14; Studio notebook tests isolate
+  configuration instead of depending on a developer's private `.env`.
+- Live CI records disabled routes as not-run, never as successful live evidence.
+
+Registration of a Trusted Publisher is not proof of an OIDC upload. Release
+acceptance still requires the actual workflow runs and new exact-wheel evidence.
+
 ## 2.1.1 - 2026-08-30
 
 Patch release that keeps the 2.1 grammar stable while aligning the published

{agentic_systems-2.1.1/src/agentic_systems.egg-info → agentic_systems-2.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentic-systems
-Version: 2.1.1
+Version: 2.1.2
 Summary: Agentic Systems: a computational grammar for building, executing, observing, and evaluating intelligent systems.
 Author-email: Jacobo Gerardo Gonzalez Leon <jacobogerardo.gonzalez@bbva.com>
 License-Expression: MIT
@@ -16,7 +16,7 @@ Classifier: Programming Language :: Python :: 3.14
 Classifier: Operating System :: OS Independent
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
-Requires-Dist: pydantic>=2.7.0
+Requires-Dist: pydantic<3,>=2.7.0
 Requires-Dist: pandas>=2.0.0
 Requires-Dist: typing-extensions>=4.10.0
 Requires-Dist: rich>=13.0.0
@@ -87,10 +87,9 @@ Requires-Dist: tomli>=2.1.0; python_version < "3.11" and extra == "all"
 </p>
 
 <p align="center">
-  <a href="https://github.com/JacoboGGLeon/agentic_systems/releases/tag/v2.1.1"><img src="https://img.shields.io/badge/release-v2.1.1-blue.svg" alt="Agentic Systems 2.1.1 release" /></a>
+  <a href="https://pypi.org/project/agentic-systems/"><img src="https://img.shields.io/pypi/v/agentic-systems" alt="Latest published Agentic Systems version" /></a>
   <a href="INSTALL.md"><img src="https://img.shields.io/badge/python-%3E%3D3.10-blue.svg" alt="Python >=3.10" /></a>
-  <img src="https://img.shields.io/badge/core%20coverage-100%25-brightgreen.svg" alt="Core coverage 100%; Bedrock separately gated" />
-  <img src="https://img.shields.io/badge/tests-passing-brightgreen.svg" alt="Tests passing" />
+  <a href="https://github.com/JacoboGGLeon/agentic_systems/actions/workflows/quality.yml"><img src="https://github.com/JacoboGGLeon/agentic_systems/actions/workflows/quality.yml/badge.svg" alt="Quality and coverage gates" /></a>
 </p>
 
 **Agentic Systems is a computational grammar for building intelligent systems
@@ -129,31 +128,46 @@ surface must explain the same product without overstating its evidence.
 
 Public concepts are defined in the API, explained in the documentation, taught through the canonical tutorials, and checked by explicit release gates.
 
-Release status: `2.1.1` defines a canonical 5 Provider x 4 Framework registry,
+API target: `2.1.2` defines a canonical 5 Provider x 4 Framework registry,
 strict versioned schemas, normalized reasoning-safe results and polymorphic
 adapter contracts. Its public surface contains 89 stable top-level exports and
 467 traced export/member IDs; all 2.0 serialization views remain compatible.
 
-The certified release passed the 20-route primary Provider x Framework matrix
+The published `2.1.1` certification passed the 20-route primary Provider x Framework matrix
 and 76/76 primary semantic episodes across deterministic Python, OpenAI,
 Ollama, Bedrock API key and vLLM. The same checksum-verified wheel also passed
 Bedrock's four-framework AWS credential-chain matrix in AWS SageMaker and in
 the ADA enterprise sandbox: 28/28 certified routes and 108/108 reviewed
 semantic episodes in total. These are artifact-specific results, not a claim
 that every credential, endpoint or model is available in every environment.
+They do not certify a different wheel. Each subsequent release must supply its
+own exact-wheel certification summary and publication evidence.
+
+The `2.1.2` release acceptance profile requires OpenAI, Bedrock API key and
+Ollama across native, LangGraph, OpenAI Agents and Strands (12 live combinations),
+plus a four-framework deterministic Python control. Each route is checked in
+two independent passes against the exact wheel. The matching certification
+summary records results; this declaration alone is not evidence of a pass.
+vLLM on Colab GPU, Bedrock IAM on AWS and Bedrock IAM in ADA remain pending
+for this version. ADA is a deployment environment, not another Provider.
 
 ## Installation
 
-Install the computational grammar from PyPI:
+The examples below target `2.1.2`. A development checkout may precede publication:
+check the linked PyPI version before installing, or use the supplied candidate
+wheel when validating a prerelease. Final release links become available only
+after certification and publication.
+
+Install the computational grammar from PyPI once that version is published:
 
 ```bash
-python -m pip install "agentic-systems==2.1.1"
+python -m pip install "agentic-systems==2.1.2"
 ```
 
 OpenAI Agents is optional and uses a different distribution/import name:
 
 ```bash
-python -m pip install "agentic-systems[openai-agents]==2.1.1"
+python -m pip install "agentic-systems[openai-agents]==2.1.2"
 # Dependency distribution: openai-agents; Python import: agents
 ```
 
@@ -161,7 +175,7 @@ For the complete portable Provider/Framework tutorial stack (excluding the
 platform-specific vLLM GPU server):
 
 ```bash
-python -m pip install "agentic-systems[all]==2.1.1"
+python -m pip install "agentic-systems[all]==2.1.2"
 ```
 
 ```python
@@ -171,39 +185,43 @@ import agentic_systems as toolkit
 The PyPI distribution contains the runtime, public API and CLI. Studio and its
 notebooks are a separate portable application so they can carry their own
 `.env.example`, tests and presentation dependencies without inflating the core
-wheel. Download `agentic-systems-studio-2.1.1.zip` from the matching GitHub
+wheel. Download `agentic-systems-studio-2.1.2.zip` from the matching GitHub
 release, then install its `ui,notebook` extras as described in the included
 README.
 
 ### Install the Agentic Systems skill
 
-The `v2.1.1` GitHub release includes a credential-free skill ZIP whose archive
+The `v2.1.2` delivery includes a credential-free skill ZIP whose archive
 root is `agentic-systems/`.
 
 ```powershell
-Expand-Archive .\agentic-systems-skill-2.1.1.zip `
+Expand-Archive .\agentic-systems-skill-2.1.2.zip `
   -DestinationPath "$env:USERPROFILE\.codex\skills" -Force
 ```
 
 ```bash
-unzip agentic-systems-skill-2.1.1.zip -d ~/.codex/skills
+unzip agentic-systems-skill-2.1.2.zip -d ~/.codex/skills
 ```
 
 Restart or reload Codex, then invoke `$agentic-systems`. OpenAI Skills upload
 surfaces that accept a ZIP can consume the same artifact. Download the
-standalone skill, conversational Studio and ADA offline bundle from the
-[v2.1.1 release](https://github.com/JacoboGGLeon/agentic_systems/releases/tag/v2.1.1).
+standalone skill and conversational Studio from the
+[v2.1.2 release](https://github.com/JacoboGGLeon/agentic_systems/releases/tag/v2.1.2).
 
 Release assets:
 
-- `agentic_systems-2.1.1-py3-none-any.whl`: certified universal Python wheel.
-- `agentic_systems-2.1.1.tar.gz`: matching source distribution.
-- `agentic-systems-skill-2.1.1.zip`: Codex/OpenAI skill package.
-- `agentic-systems-studio-2.1.1.zip`: one portable conversational Agentic System.
-- `agentic-systems-2.1.1-ada-offline.zip`: wheel, Studio, tutorials and live evidence for restricted environments.
-- `agentic-systems-2.1.1-strands-protocol-challenge.zip`: frozen MCP+A2A/LangGraph semantic gate for local, Colab, SageMaker and ADA execution.
+- `agentic_systems-2.1.2-py3-none-any.whl`: universal Python wheel; certification is recorded in the matching summary.
+- `agentic_systems-2.1.2.tar.gz`: matching source distribution.
+- `agentic-systems-skill-2.1.2.zip`: Codex/OpenAI skill package.
+- `agentic-systems-studio-2.1.2.zip`: one portable conversational Agentic System.
+- `agentic-systems-2.1.2-strands-protocol-challenge.zip`: frozen MCP+A2A/LangGraph semantic gate for local, Colab, SageMaker and ADA execution.
 - `final-certification-summary.json`: exact wheel identity and reviewed live evidence inventory.
-- `SHA256SUMS-2.1.1.txt`: hashes for the Python and product artifacts.
+- `SHA256SUMS-2.1.2.txt`: hashes for the Python and product artifacts.
+- `release-manifest.json`: typed artifact inventory used to verify and repeat publication without replacing bytes.
+
+ADA deployment bundles are delivered separately after validation in the target
+environment. There is no `[ada]` extra in 2.1.2. Bedrock IAM uses the existing
+`bedrock` extra; installing dependencies does not certify an execution route.
 
 For an installed-package smoke test and provider notebook setup, follow
 [First Run Onboarding](docs/ONBOARDING_FIRST_RUN.md).
@@ -1009,7 +1027,7 @@ evidence belongs to GitHub Releases.
 Current verified status:
 
 ```text
-Version: 2.1.1
+Version: 2.1.2
 PyPI package: agentic-systems
 Tests: run `python -m pytest` for the current count
 Core coverage: 100.00%
@@ -1034,6 +1052,10 @@ python -m compileall -q src tests tutorials
 The complete production, Pydantic and POO/polymorphism certification is
 documented in [Triple Quality Gate](docs/QUALITY_GATES.md).
 
+Maintainers publish through the [Trusted Publishing runbook](docs/RELEASE_PUBLISHING.md):
+one certified artifact set, explicit hash checks, and repeatable publication
+without replacing an existing release.
+
 For full coverage validation:
 
 ```bash

{agentic_systems-2.1.1 → agentic_systems-2.1.2}/README.md

@@ -5,10 +5,9 @@
 </p>
 
 <p align="center">
-  <a href="https://github.com/JacoboGGLeon/agentic_systems/releases/tag/v2.1.1"><img src="https://img.shields.io/badge/release-v2.1.1-blue.svg" alt="Agentic Systems 2.1.1 release" /></a>
+  <a href="https://pypi.org/project/agentic-systems/"><img src="https://img.shields.io/pypi/v/agentic-systems" alt="Latest published Agentic Systems version" /></a>
   <a href="INSTALL.md"><img src="https://img.shields.io/badge/python-%3E%3D3.10-blue.svg" alt="Python >=3.10" /></a>
-  <img src="https://img.shields.io/badge/core%20coverage-100%25-brightgreen.svg" alt="Core coverage 100%; Bedrock separately gated" />
-  <img src="https://img.shields.io/badge/tests-passing-brightgreen.svg" alt="Tests passing" />
+  <a href="https://github.com/JacoboGGLeon/agentic_systems/actions/workflows/quality.yml"><img src="https://github.com/JacoboGGLeon/agentic_systems/actions/workflows/quality.yml/badge.svg" alt="Quality and coverage gates" /></a>
 </p>
 
 **Agentic Systems is a computational grammar for building intelligent systems
@@ -47,31 +46,46 @@ surface must explain the same product without overstating its evidence.
 
 Public concepts are defined in the API, explained in the documentation, taught through the canonical tutorials, and checked by explicit release gates.
 
-Release status: `2.1.1` defines a canonical 5 Provider x 4 Framework registry,
+API target: `2.1.2` defines a canonical 5 Provider x 4 Framework registry,
 strict versioned schemas, normalized reasoning-safe results and polymorphic
 adapter contracts. Its public surface contains 89 stable top-level exports and
 467 traced export/member IDs; all 2.0 serialization views remain compatible.
 
-The certified release passed the 20-route primary Provider x Framework matrix
+The published `2.1.1` certification passed the 20-route primary Provider x Framework matrix
 and 76/76 primary semantic episodes across deterministic Python, OpenAI,
 Ollama, Bedrock API key and vLLM. The same checksum-verified wheel also passed
 Bedrock's four-framework AWS credential-chain matrix in AWS SageMaker and in
 the ADA enterprise sandbox: 28/28 certified routes and 108/108 reviewed
 semantic episodes in total. These are artifact-specific results, not a claim
 that every credential, endpoint or model is available in every environment.
+They do not certify a different wheel. Each subsequent release must supply its
+own exact-wheel certification summary and publication evidence.
+
+The `2.1.2` release acceptance profile requires OpenAI, Bedrock API key and
+Ollama across native, LangGraph, OpenAI Agents and Strands (12 live combinations),
+plus a four-framework deterministic Python control. Each route is checked in
+two independent passes against the exact wheel. The matching certification
+summary records results; this declaration alone is not evidence of a pass.
+vLLM on Colab GPU, Bedrock IAM on AWS and Bedrock IAM in ADA remain pending
+for this version. ADA is a deployment environment, not another Provider.
 
 ## Installation
 
-Install the computational grammar from PyPI:
+The examples below target `2.1.2`. A development checkout may precede publication:
+check the linked PyPI version before installing, or use the supplied candidate
+wheel when validating a prerelease. Final release links become available only
+after certification and publication.
+
+Install the computational grammar from PyPI once that version is published:
 
 ```bash
-python -m pip install "agentic-systems==2.1.1"
+python -m pip install "agentic-systems==2.1.2"
 ```
 
 OpenAI Agents is optional and uses a different distribution/import name:
 
 ```bash
-python -m pip install "agentic-systems[openai-agents]==2.1.1"
+python -m pip install "agentic-systems[openai-agents]==2.1.2"
 # Dependency distribution: openai-agents; Python import: agents
 ```
 
@@ -79,7 +93,7 @@ For the complete portable Provider/Framework tutorial stack (excluding the
 platform-specific vLLM GPU server):
 
 ```bash
-python -m pip install "agentic-systems[all]==2.1.1"
+python -m pip install "agentic-systems[all]==2.1.2"
 ```
 
 ```python
@@ -89,39 +103,43 @@ import agentic_systems as toolkit
 The PyPI distribution contains the runtime, public API and CLI. Studio and its
 notebooks are a separate portable application so they can carry their own
 `.env.example`, tests and presentation dependencies without inflating the core
-wheel. Download `agentic-systems-studio-2.1.1.zip` from the matching GitHub
+wheel. Download `agentic-systems-studio-2.1.2.zip` from the matching GitHub
 release, then install its `ui,notebook` extras as described in the included
 README.
 
 ### Install the Agentic Systems skill
 
-The `v2.1.1` GitHub release includes a credential-free skill ZIP whose archive
+The `v2.1.2` delivery includes a credential-free skill ZIP whose archive
 root is `agentic-systems/`.
 
 ```powershell
-Expand-Archive .\agentic-systems-skill-2.1.1.zip `
+Expand-Archive .\agentic-systems-skill-2.1.2.zip `
   -DestinationPath "$env:USERPROFILE\.codex\skills" -Force
 ```
 
 ```bash
-unzip agentic-systems-skill-2.1.1.zip -d ~/.codex/skills
+unzip agentic-systems-skill-2.1.2.zip -d ~/.codex/skills
 ```
 
 Restart or reload Codex, then invoke `$agentic-systems`. OpenAI Skills upload
 surfaces that accept a ZIP can consume the same artifact. Download the
-standalone skill, conversational Studio and ADA offline bundle from the
-[v2.1.1 release](https://github.com/JacoboGGLeon/agentic_systems/releases/tag/v2.1.1).
+standalone skill and conversational Studio from the
+[v2.1.2 release](https://github.com/JacoboGGLeon/agentic_systems/releases/tag/v2.1.2).
 
 Release assets:
 
-- `agentic_systems-2.1.1-py3-none-any.whl`: certified universal Python wheel.
-- `agentic_systems-2.1.1.tar.gz`: matching source distribution.
-- `agentic-systems-skill-2.1.1.zip`: Codex/OpenAI skill package.
-- `agentic-systems-studio-2.1.1.zip`: one portable conversational Agentic System.
-- `agentic-systems-2.1.1-ada-offline.zip`: wheel, Studio, tutorials and live evidence for restricted environments.
-- `agentic-systems-2.1.1-strands-protocol-challenge.zip`: frozen MCP+A2A/LangGraph semantic gate for local, Colab, SageMaker and ADA execution.
+- `agentic_systems-2.1.2-py3-none-any.whl`: universal Python wheel; certification is recorded in the matching summary.
+- `agentic_systems-2.1.2.tar.gz`: matching source distribution.
+- `agentic-systems-skill-2.1.2.zip`: Codex/OpenAI skill package.
+- `agentic-systems-studio-2.1.2.zip`: one portable conversational Agentic System.
+- `agentic-systems-2.1.2-strands-protocol-challenge.zip`: frozen MCP+A2A/LangGraph semantic gate for local, Colab, SageMaker and ADA execution.
 - `final-certification-summary.json`: exact wheel identity and reviewed live evidence inventory.
-- `SHA256SUMS-2.1.1.txt`: hashes for the Python and product artifacts.
+- `SHA256SUMS-2.1.2.txt`: hashes for the Python and product artifacts.
+- `release-manifest.json`: typed artifact inventory used to verify and repeat publication without replacing bytes.
+
+ADA deployment bundles are delivered separately after validation in the target
+environment. There is no `[ada]` extra in 2.1.2. Bedrock IAM uses the existing
+`bedrock` extra; installing dependencies does not certify an execution route.
 
 For an installed-package smoke test and provider notebook setup, follow
 [First Run Onboarding](docs/ONBOARDING_FIRST_RUN.md).
@@ -927,7 +945,7 @@ evidence belongs to GitHub Releases.
 Current verified status:
 
 ```text
-Version: 2.1.1
+Version: 2.1.2
 PyPI package: agentic-systems
 Tests: run `python -m pytest` for the current count
 Core coverage: 100.00%
@@ -950,7 +968,11 @@ python -m compileall -q src tests tutorials
 ```
 
 The complete production, Pydantic and POO/polymorphism certification is
-documented in [Triple Quality Gate](docs/QUALITY_GATES.md).
+documented in [Triple Quality Gate](docs/QUALITY_GATES.md).
+
+Maintainers publish through the [Trusted Publishing runbook](docs/RELEASE_PUBLISHING.md):
+one certified artifact set, explicit hash checks, and repeatable publication
+without replacing an existing release.
 
 For full coverage validation:
 

{agentic_systems-2.1.1 → agentic_systems-2.1.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "agentic-systems"
-version = "2.1.1"
+version = "2.1.2"
 description = "Agentic Systems: a computational grammar for building, executing, observing, and evaluating intelligent systems."
 readme = "README.md"
 requires-python = ">=3.10"
@@ -20,7 +20,7 @@ classifiers = [
     "Operating System :: OS Independent",
 ]
 dependencies = [
-    "pydantic>=2.7.0",
+    "pydantic>=2.7.0,<3",
     "pandas>=2.0.0",
     "typing-extensions>=4.10.0",
     "rich>=13.0.0",

{agentic_systems-2.1.1 → agentic_systems-2.1.2}/src/agentic_systems/__init__.py

@@ -112,7 +112,7 @@ from .execution import (
     coerce_run_result,
     is_executable,
 )
-from .lineage import LINEAGE_SCHEMA_VERSION, LineageMemory, LineageStep, lineage_memory
+from .lineage import LINEAGE_SCHEMA_VERSION, LineageMemory, LineageStep, lineage_memory
 from .final_answer import (
     FINAL_ANSWER_SCHEMA_VERSION,
     OutputSchema,
@@ -162,7 +162,7 @@ from .providers.openai_runtime import openai_environment_snapshot
 from .providers.ollama_runtime import ollama_environment_snapshot
 from .providers.vllm_runtime import vllm_environment_snapshot
 
-__version__ = "2.1.1"
+__version__ = "2.1.2"
 
 __all__ = list(_PUBLIC_API)
 del _PUBLIC_API