matprops-sdk 0.1.1__tar.gz

matprops_sdk-0.1.1/.gitignore

@@ -0,0 +1,60 @@
+.DS_Store
+node_modules/
+**/node_modules/
+.build/
+**/.build/
+*.tsbuildinfo
+.next/
+**/.next/
+.env
+.env.*
+!.env.sample*
+# Package-level names are symlinks to the root files (git stores the link, not secrets)
+!/frontend/.env
+!/frontend/.env.local
+!/db/.env
+!/db/.env.local
+*.log
+__pycache__/
+*.pyc
+.venv/
+venv/
+dist/
+build/
+.turbo/
+.serverless/
+.webpack/
+coverage/
+.nyc_output/
+*.egg-info/
+.pytest_cache/
+.ruff_cache/
+.run-results/
+matprops-run/
+results/
+.pnpm-store/
+.pulumi/
+.vercel/
+
+.envy.*
+
+# lattice-dynamics generated artifacts
+mesh_dicts/
+mesh_dicts*/
+outfiles/
+outfiles*/
+kappa-m*
+*IFCs/
+fc*
+FORCE_SETS
+FORCES_FC2
+FORCES_FC3
+POSCAR*
+BORN
+aux_scripts/
+gruneisen*/
+lattice-dynamics/**/*output/
+
+# Local-only reference code — never commit or push
+/example-codebase/
+example-codebase/**

matprops_sdk-0.1.1/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 MatProps
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

matprops_sdk-0.1.1/PKG-INFO

@@ -0,0 +1,165 @@
+Metadata-Version: 2.5
+Name: matprops-sdk
+Version: 0.1.1
+Summary: Define and run Matprops simulations locally.
+Project-URL: Homepage, https://matprops.com
+Project-URL: Repository, https://github.com/rileyhanus/matprops
+Project-URL: Issues, https://github.com/rileyhanus/matprops/issues
+Author: MatProps
+License-Expression: MIT
+License-File: LICENSE
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering
+Requires-Python: >=3.11
+Requires-Dist: pydantic<3,>=2
+Provides-Extra: dev
+Requires-Dist: pytest>=8; extra == 'dev'
+Requires-Dist: ruff; extra == 'dev'
+Description-Content-Type: text/markdown
+
+# Matprops SDK
+
+Local toolkit for typed material simulations: validate the contract, emit JSON Schema, run on your machine, and package or publish a zip to sim-server.
+
+Requires Python 3.11+.
+
+```bash
+python -m pip install matprops-sdk
+matprops init my-sim
+cd my-sim
+matprops validate
+matprops run --input example.input.json
+matprops package            # dist/<name>.zip, upload it in the web UI
+matprops publish            # or package + upload in one step (needs login)
+```
+
+From a clone of this repo, install an editable copy with tests and lint:
+
+```bash
+python -m pip install -e ".[dev]"
+```
+
+`init` writes `simulation.py`, `pyproject.toml`, and `example.input.json` (it will not overwrite existing files). Entrypoint defaults to `[tool.matprops] entrypoint = "simulation:run"`; override with `--entrypoint`.
+
+## CLI login
+
+`matprops login` opens the Matprops web app (or prints a URL) so you can sign in and approve the CLI. Temporary credentials are stored in `~/.config/matprops/credentials.json` (`chmod 600`).
+
+```bash
+matprops login
+matprops whoami
+matprops logout
+```
+
+Defaults are production: `https://matprops.com` and `https://api.matprops.com`. Override with `MATPROPS_APP_URL` / `MATPROPS_API_URL` or `--app-url` / `--api-url`. For the local stack use `http://localhost:3000` and `http://localhost:4000`; for the shared dev environment use `https://dev.matprops.com` and `https://api.dev.matprops.com`. Use `--no-browser` on a remote machine and open the printed URL locally.
+
+## Simulation
+
+Use Pydantic models and `@simulation`. Sibling modules are importable as long as filenames are valid Python (`my_logic.py`, not `my-logic.py`).
+
+```python
+from matprops_sdk import Field, InputModel, OutputModel, RunContext, simulation
+
+
+class Inputs(InputModel):
+    radius: float = Field(gt=0, units="cm")
+
+
+class Outputs(OutputModel):
+    volume: float = Field(units="cm^3")
+
+
+@simulation(name="volume-of-sphere", inputs=Inputs, outputs=Outputs)
+def run(inputs: Inputs, context: RunContext) -> Outputs:
+    path = context.artifact("trace.txt", "Calculation trace")
+    path.write_text("...", encoding="utf-8")
+    return Outputs(volume=...)
+```
+
+The kebab-case `name` is the compute identity. Optional decorator metadata (`cpu`, `memory`, `timeout`) is recorded in schema/manifest and not enforced locally.
+
+`validate` checks load, decorator, signature, and schema only — not physics and not `example.input.json`.
+
+Walkthrough with split logic: [`examples/volume-of-sphere/`](examples/volume-of-sphere/).
+
+## CLI
+
+| Command | What it does |
+| --- | --- |
+| `matprops init [dir]` | Scaffold a project |
+| `matprops validate` | Check the entrypoint contract |
+| `matprops schema [--output-dir DIR]` | Print schemas, or write `input.schema.json` / `output.schema.json` |
+| `matprops run --input …` | Run locally |
+| `matprops login` | Sign in via the web app and store credentials |
+| `matprops logout` | Revoke and delete stored credentials |
+| `matprops whoami` | Show the current CLI identity |
+| `matprops package [--output PATH]` | Validate and write an uploadable zip to `dist/<name>.zip` |
+| `matprops publish` | Package the project and POST it to sim-server |
+
+`--input` is a file path, or inline JSON if the trimmed value starts with `{` or `[`:
+
+```bash
+matprops run --input '{"var1": 1, "var2": 2}'
+```
+
+Each run writes a timestamped folder under `.run-results/` (UTC, millisecond stamp, e.g. `.run-results/2026-09-01T15-11-22-123Z/`) with `outputs.json`, `manifest.json`, and `artifacts/`. `--output-dir` changes that parent directory.
+
+Add `.run-results/` to gitignore.
+
+## Packaging and publishing
+
+`package` validates the contract in your environment (where your `requirements.txt` dependencies are installed), then zips the project directory together with a generated `manifest.json` holding the simulation name, entrypoint, input/output JSON Schema and runtime hints. It skips `.git`, virtualenvs, `.run-results`, `dist`, caches, `*.pyc`, `.DS_Store` and root-level `example.input*.json` files. The zip is written to `dist/<name>.zip` (or `--output`) and is byte-for-byte reproducible for identical code.
+
+The platform only accepts zips that contain this manifest: it reads the contract from `manifest.json` instead of importing your code, so uploads are validated instantly regardless of which packages the simulation needs. Upload the zip in the web UI, or let `publish` do it for you. Either way the image build installs your `requirements.txt` and re-derives the contract to confirm the manifest.
+
+`publish` runs the same packaging step and POSTs the bundle to `{MATPROPS_API_URL}/simulation-function-versions` (default `https://api.matprops.com`). Authenticate with `matprops login`, `MATPROPS_API_KEY`, or `--api-key`. Override the URL with `MATPROPS_API_URL` or `--url`.
+
+Add `dist/` to gitignore.
+
+## Releasing
+
+PyPI project name is `matprops-sdk` (`matprops` is taken). Versions live in both `pyproject.toml` (`[project].version`) and `src/matprops_sdk/__init__.py` (`__version__`); bump them together.
+
+Do **not** tag releases as `v1`, `v13`, or other `v*` integers — those tags are used by the production deploy workflow. SDK tags are `sdk-vX.Y.Z` only.
+
+### One-time setup
+
+1. Create a [pypi.org](https://pypi.org/account/register) account with 2FA (and optionally a separate [test.pypi.org](https://test.pypi.org) account).
+2. In GitHub: **Settings → Environments** → create `pypi` (URL `https://pypi.org/project/matprops-sdk/`) and optionally `testpypi`. Required reviewers are recommended on `pypi`.
+3. On PyPI, add a [pending trusted publisher](https://pypi.org/manage/account/publishing/):
+   - Project name: `matprops-sdk`
+   - Owner: `rileyhanus`
+   - Repository: `matprops`
+   - Workflow filename: `python-sdk-publish.yml`
+   - Environment name: `pypi`
+4. Repeat on TestPyPI with environment `testpypi` if you want a dry run.
+
+A typo in the workflow name or environment is a silent 403 on first upload.
+
+### Publish a version
+
+1. Bump both version fields, merge to `main`.
+2. From `python/matprops-sdk`: `pip install -e ".[dev]" build twine`, then `ruff check . && pytest`, `python -m build`, and `python -m twine check dist/*`.
+3. Tag the merged main SHA and push:
+
+   ```bash
+   git tag -a sdk-v0.1.0 -m "matprops-sdk 0.1.0"
+   git push origin sdk-v0.1.0
+   ```
+
+4. Watch **Actions → Publish Python SDK**. Approve the `pypi` environment if reviewers are required.
+
+The first successful trusted-publisher run creates `https://pypi.org/project/matprops-sdk/` and uploads that version. PyPI versions are immutable; a bad upload means yank and ship the next version.
+
+Optional TestPyPI dry run: **Actions → Publish Python SDK → Run workflow** with target `testpypi`, then:
+
+```bash
+python -m pip install -i https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple/ matprops-sdk
+```
+
+A first upload can also be done from a laptop with an account-scoped API token (`python -m twine upload dist/*`; username `__token__`). Later releases should use the `sdk-v*` tag so CI and Trusted Publishing stay the path of record.

matprops_sdk-0.1.1/README.md

@@ -0,0 +1,141 @@
+# Matprops SDK
+
+Local toolkit for typed material simulations: validate the contract, emit JSON Schema, run on your machine, and package or publish a zip to sim-server.
+
+Requires Python 3.11+.
+
+```bash
+python -m pip install matprops-sdk
+matprops init my-sim
+cd my-sim
+matprops validate
+matprops run --input example.input.json
+matprops package            # dist/<name>.zip, upload it in the web UI
+matprops publish            # or package + upload in one step (needs login)
+```
+
+From a clone of this repo, install an editable copy with tests and lint:
+
+```bash
+python -m pip install -e ".[dev]"
+```
+
+`init` writes `simulation.py`, `pyproject.toml`, and `example.input.json` (it will not overwrite existing files). Entrypoint defaults to `[tool.matprops] entrypoint = "simulation:run"`; override with `--entrypoint`.
+
+## CLI login
+
+`matprops login` opens the Matprops web app (or prints a URL) so you can sign in and approve the CLI. Temporary credentials are stored in `~/.config/matprops/credentials.json` (`chmod 600`).
+
+```bash
+matprops login
+matprops whoami
+matprops logout
+```
+
+Defaults are production: `https://matprops.com` and `https://api.matprops.com`. Override with `MATPROPS_APP_URL` / `MATPROPS_API_URL` or `--app-url` / `--api-url`. For the local stack use `http://localhost:3000` and `http://localhost:4000`; for the shared dev environment use `https://dev.matprops.com` and `https://api.dev.matprops.com`. Use `--no-browser` on a remote machine and open the printed URL locally.
+
+## Simulation
+
+Use Pydantic models and `@simulation`. Sibling modules are importable as long as filenames are valid Python (`my_logic.py`, not `my-logic.py`).
+
+```python
+from matprops_sdk import Field, InputModel, OutputModel, RunContext, simulation
+
+
+class Inputs(InputModel):
+    radius: float = Field(gt=0, units="cm")
+
+
+class Outputs(OutputModel):
+    volume: float = Field(units="cm^3")
+
+
+@simulation(name="volume-of-sphere", inputs=Inputs, outputs=Outputs)
+def run(inputs: Inputs, context: RunContext) -> Outputs:
+    path = context.artifact("trace.txt", "Calculation trace")
+    path.write_text("...", encoding="utf-8")
+    return Outputs(volume=...)
+```
+
+The kebab-case `name` is the compute identity. Optional decorator metadata (`cpu`, `memory`, `timeout`) is recorded in schema/manifest and not enforced locally.
+
+`validate` checks load, decorator, signature, and schema only — not physics and not `example.input.json`.
+
+Walkthrough with split logic: [`examples/volume-of-sphere/`](examples/volume-of-sphere/).
+
+## CLI
+
+| Command | What it does |
+| --- | --- |
+| `matprops init [dir]` | Scaffold a project |
+| `matprops validate` | Check the entrypoint contract |
+| `matprops schema [--output-dir DIR]` | Print schemas, or write `input.schema.json` / `output.schema.json` |
+| `matprops run --input …` | Run locally |
+| `matprops login` | Sign in via the web app and store credentials |
+| `matprops logout` | Revoke and delete stored credentials |
+| `matprops whoami` | Show the current CLI identity |
+| `matprops package [--output PATH]` | Validate and write an uploadable zip to `dist/<name>.zip` |
+| `matprops publish` | Package the project and POST it to sim-server |
+
+`--input` is a file path, or inline JSON if the trimmed value starts with `{` or `[`:
+
+```bash
+matprops run --input '{"var1": 1, "var2": 2}'
+```
+
+Each run writes a timestamped folder under `.run-results/` (UTC, millisecond stamp, e.g. `.run-results/2026-09-01T15-11-22-123Z/`) with `outputs.json`, `manifest.json`, and `artifacts/`. `--output-dir` changes that parent directory.
+
+Add `.run-results/` to gitignore.
+
+## Packaging and publishing
+
+`package` validates the contract in your environment (where your `requirements.txt` dependencies are installed), then zips the project directory together with a generated `manifest.json` holding the simulation name, entrypoint, input/output JSON Schema and runtime hints. It skips `.git`, virtualenvs, `.run-results`, `dist`, caches, `*.pyc`, `.DS_Store` and root-level `example.input*.json` files. The zip is written to `dist/<name>.zip` (or `--output`) and is byte-for-byte reproducible for identical code.
+
+The platform only accepts zips that contain this manifest: it reads the contract from `manifest.json` instead of importing your code, so uploads are validated instantly regardless of which packages the simulation needs. Upload the zip in the web UI, or let `publish` do it for you. Either way the image build installs your `requirements.txt` and re-derives the contract to confirm the manifest.
+
+`publish` runs the same packaging step and POSTs the bundle to `{MATPROPS_API_URL}/simulation-function-versions` (default `https://api.matprops.com`). Authenticate with `matprops login`, `MATPROPS_API_KEY`, or `--api-key`. Override the URL with `MATPROPS_API_URL` or `--url`.
+
+Add `dist/` to gitignore.
+
+## Releasing
+
+PyPI project name is `matprops-sdk` (`matprops` is taken). Versions live in both `pyproject.toml` (`[project].version`) and `src/matprops_sdk/__init__.py` (`__version__`); bump them together.
+
+Do **not** tag releases as `v1`, `v13`, or other `v*` integers — those tags are used by the production deploy workflow. SDK tags are `sdk-vX.Y.Z` only.
+
+### One-time setup
+
+1. Create a [pypi.org](https://pypi.org/account/register) account with 2FA (and optionally a separate [test.pypi.org](https://test.pypi.org) account).
+2. In GitHub: **Settings → Environments** → create `pypi` (URL `https://pypi.org/project/matprops-sdk/`) and optionally `testpypi`. Required reviewers are recommended on `pypi`.
+3. On PyPI, add a [pending trusted publisher](https://pypi.org/manage/account/publishing/):
+   - Project name: `matprops-sdk`
+   - Owner: `rileyhanus`
+   - Repository: `matprops`
+   - Workflow filename: `python-sdk-publish.yml`
+   - Environment name: `pypi`
+4. Repeat on TestPyPI with environment `testpypi` if you want a dry run.
+
+A typo in the workflow name or environment is a silent 403 on first upload.
+
+### Publish a version
+
+1. Bump both version fields, merge to `main`.
+2. From `python/matprops-sdk`: `pip install -e ".[dev]" build twine`, then `ruff check . && pytest`, `python -m build`, and `python -m twine check dist/*`.
+3. Tag the merged main SHA and push:
+
+   ```bash
+   git tag -a sdk-v0.1.0 -m "matprops-sdk 0.1.0"
+   git push origin sdk-v0.1.0
+   ```
+
+4. Watch **Actions → Publish Python SDK**. Approve the `pypi` environment if reviewers are required.
+
+The first successful trusted-publisher run creates `https://pypi.org/project/matprops-sdk/` and uploads that version. PyPI versions are immutable; a bad upload means yank and ship the next version.
+
+Optional TestPyPI dry run: **Actions → Publish Python SDK → Run workflow** with target `testpypi`, then:
+
+```bash
+python -m pip install -i https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple/ matprops-sdk
+```
+
+A first upload can also be done from a laptop with an account-scoped API token (`python -m twine upload dist/*`; username `__token__`). Later releases should use the `sdk-v*` tag so CI and Trusted Publishing stay the path of record.

matprops_sdk-0.1.1/examples/add_numbers/example.input.json

@@ -0,0 +1 @@
+{"num1": 1.5, "num2": 2.5}

matprops_sdk-0.1.1/examples/add_numbers/pyproject.toml

@@ -0,0 +1,8 @@
+[project]
+name = "matprops-example-add-numbers"
+version = "0.1.0"
+requires-python = ">=3.11"
+dependencies = ["matprops-sdk"]
+
+[tool.matprops]
+entrypoint = "simulation:run"

matprops_sdk-0.1.1/examples/add_numbers/simulation.py

@@ -0,0 +1,15 @@
+from matprops_sdk import Field, InputModel, OutputModel, simulation
+
+
+class Inputs(InputModel):
+    num1: float = Field(description="First number", units="unitless")
+    num2: float = Field(description="Second number", units="unitless")
+
+
+class Outputs(OutputModel):
+    sum: float = Field(description="Sum of the two numbers", units="unitless")
+
+
+@simulation(name="add-numbers", inputs=Inputs, outputs=Outputs)
+def run(inputs: Inputs) -> Outputs:
+    return Outputs(sum=inputs.num1 + inputs.num2)

matprops_sdk-0.1.1/examples/composite-thermal-conductivity/example.input.json

@@ -0,0 +1,8 @@
+{
+  "k_matrix": 0.2,
+  "k_inclusion": 30.0,
+  "volume_fraction": 0.3,
+  "grid_size": 800,
+  "max_iterations": 60000,
+  "seed": 42
+}

matprops_sdk-0.1.1/examples/composite-thermal-conductivity/example.input.smoke.json

@@ -0,0 +1,8 @@
+{
+  "k_matrix": 0.2,
+  "k_inclusion": 30.0,
+  "volume_fraction": 0.3,
+  "grid_size": 32,
+  "max_iterations": 50,
+  "seed": 42
+}

matprops_sdk-0.1.1/examples/composite-thermal-conductivity/pyproject.toml

@@ -0,0 +1,8 @@
+[project]
+name = "matprops-example-composite-thermal-conductivity"
+version = "0.1.0"
+requires-python = ">=3.11"
+dependencies = ["matprops-sdk"]
+
+[tool.matprops]
+entrypoint = "simulation:run"

matprops_sdk-0.1.1/examples/composite-thermal-conductivity/requirements.txt

@@ -0,0 +1 @@
+numpy>=1.26

matprops_sdk-0.1.1/examples/composite-thermal-conductivity/simulation.py

@@ -0,0 +1,156 @@
+import numpy as np
+from solver import (
+    conductivity_field,
+    effective_conductivity,
+    hashin_shtrikman_2d,
+    random_microstructure,
+    solve_temperature,
+)
+
+from matprops_sdk import (
+    Field,
+    InputModel,
+    OutputModel,
+    RunContext,
+    simulation,
+)
+
+
+class Inputs(InputModel):
+    k_matrix: float = Field(
+        default=0.2,
+        gt=0,
+        units="W/m·K",
+        description="Thermal conductivity of the continuous matrix",
+    )
+    k_inclusion: float = Field(
+        default=30.0,
+        gt=0,
+        units="W/m·K",
+        description="Thermal conductivity of the dispersed inclusions",
+    )
+    volume_fraction: float = Field(
+        default=0.3,
+        ge=0,
+        le=1,
+        units="unitless",
+        description="Target inclusion volume fraction",
+    )
+    grid_size: int = Field(
+        default=800,
+        ge=8,
+        le=1024,
+        units="unitless",
+        description="Number of cells along each side of the unit square",
+    )
+    max_iterations: int = Field(
+        default=60_000,
+        ge=1,
+        le=500_000,
+        units="unitless",
+        description="Jacobi sweeps; raise this (and grid_size) to spend more compute",
+    )
+    seed: int = Field(
+        default=42,
+        ge=0,
+        units="unitless",
+        description="RNG seed for the random microstructure",
+    )
+
+
+class Outputs(OutputModel):
+    k_effective: float = Field(
+        units="W/m·K",
+        description="Effective in-plane conductivity from the solved heat flux",
+    )
+    hashin_shtrikman_lower: float = Field(
+        units="W/m·K",
+        description="2D Hashin–Shtrikman lower bound for the same mixture",
+    )
+    hashin_shtrikman_upper: float = Field(
+        units="W/m·K",
+        description="2D Hashin–Shtrikman upper bound for the same mixture",
+    )
+    volume_fraction_realized: float = Field(
+        units="unitless",
+        description="Inclusion fraction actually placed on the grid",
+    )
+    residual: float = Field(
+        units="K",
+        description="Max temperature change on the last Jacobi sweep",
+    )
+
+
+@simulation(
+    name="composite-thermal-conductivity",
+    inputs=Inputs,
+    outputs=Outputs,
+    cpu=1,
+    memory="2GiB",
+    timeout="10m",
+)
+def run(inputs: Inputs, context: RunContext) -> Outputs:
+    context.progress(0, stage="generate microstructure", stage_index=0, stage_count=3)
+
+    inclusion = random_microstructure(
+        inputs.grid_size,
+        inputs.volume_fraction,
+        inputs.seed,
+    )
+    conductivity = conductivity_field(
+        inclusion,
+        inputs.k_matrix,
+        inputs.k_inclusion,
+    )
+    volume_fraction_realized = float(inclusion.mean())
+
+    context.progress(5, stage="solve heat equation", stage_index=1, stage_count=3)
+
+    def on_progress(iteration: int, total: int, _residual: float) -> None:
+        # 5%–95% is the solver; leave the last 5% for flux and artifacts.
+        percent = 5.0 + 90.0 * iteration / total
+        context.progress(
+            percent, stage="solve heat equation", stage_index=1, stage_count=3
+        )
+
+    temperature, residual = solve_temperature(
+        conductivity,
+        inputs.max_iterations,
+        progress=on_progress,
+    )
+
+    context.progress(
+        95, stage="compute effective conductivity", stage_index=2, stage_count=3
+    )
+
+    k_effective = effective_conductivity(conductivity, temperature)
+    hs_lower, hs_upper = hashin_shtrikman_2d(
+        inputs.k_matrix,
+        inputs.k_inclusion,
+        volume_fraction_realized,
+    )
+
+    np.savetxt(
+        context.artifact("microstructure.csv", "Inclusion map (1) vs matrix (0)"),
+        inclusion.astype("int8"),
+        fmt="%d",
+        delimiter=",",
+    )
+    np.savetxt(
+        context.artifact("temperature.csv", "Solved temperature field (K, relative)"),
+        temperature,
+        fmt="%.6e",
+        delimiter=",",
+    )
+
+    context.progress(
+        100, stage="compute effective conductivity", stage_index=2, stage_count=3
+    )
+
+    return Outputs(
+        k_effective=k_effective,
+        hashin_shtrikman_lower=hs_lower,
+        hashin_shtrikman_upper=hs_upper,
+        volume_fraction_realized=volume_fraction_realized,
+        residual=residual,
+    )