v0.10.1

v0.10.1 (2026-09-29)

This release is published under the MIT License.

Bug Fixes

• Correct relative import in test_state_coupling.py (a1d056f)

---

Detailed Changes: v0.10.0...v0.10.1

v0.10.0

v0.10.0 (2026-09-29)

This release is published under the MIT License.

Features

• Add dftb_hamiltonian_gradient node to calculate TD-DFTB Hamiltonian gradient and update module exports (fc1d54a)

• Introduced dftb_hamiltonian_gradient node for computing ⟨ψᵢ|∇H|ψⱼ⟩ with TD-DFTB.

• Updated Dockerfile to standardize DFTB+ installation path.

• Documented new node in README.md and updated __init__.py exports.

---

Detailed Changes: v0.9.1...v0.10.0

v0.9.1

v0.9.1 (2026-09-23)

This release is published under the MIT License.

Bug Fixes

• Add dftbplus dependency and configure Python API in Docker (82db5a1)

• Added dftbplus to pyproject.docker and uv.override-dependencies.

• Updated Dockerfile to copy and configure dftbplus Python API into the Conda environment.

• Adjusted PYTHONPATH handling for compatibility with Simstack’s runtime behavior.

---

Detailed Changes: v0.9.0...v0.9.1

v0.9.0

v0.9.0 (2026-09-22)

This release is published under the MIT License.

Features

• Add dftb_excited_states node for TD-DFTB excited-state calculations with energy and gradient outputs (1d9aa0f)

• Implements Casida TD-DFTB using the dftbplus-python API.

• Generates detailed energy and gradient outputs for selected excited states.

• Integrates with Simstack framework and supports molecular input validation.

• Handles calculation failures with optional tolerance for robustness.

---

Detailed Changes: v0.8.0...v0.9.0

v0.8.0

v0.8.0 (2026-09-22)

This release is published under the MIT License.

Features

• Add dftb_excited_states module with support for Casida TD-DFTB and update exports and documentation (3b139d6)

---

Detailed Changes: v0.7.1...v0.8.0

Apptainer SIF

Pre-built Apptainer image (molecular_qm_dftb.sif).
On 2FA-only hosts, download with a PAT (no interactive login):
curl -L -H "Authorization: Bearer $GITHUB_TOKEN" -o molecular_qm_dftb.sif https://github.com/simstack/molecular_qm_dftb/releases/download/container-sif/molecular_qm_dftb.sif

v0.7.1

v0.7.1 (2026-08-30)

This release is published under the MIT License.

Bug Fixes

• Fixhing chart_generation in dftb_calculator.py (66c4467)

---

Detailed Changes: v0.7.0...v0.7.1

v0.7.0

v0.7.0 (2026-08-28)

This release is published under the MIT License.

Features

• Add support for multiple geometry optimization methods (steepest descent, conjugate gradient, FIRE) and refactor optimization logic (006d7e1)

• Introduced OptimizationMethod enum in DftbInput for method selection.

• Added tests for new optimization methods.

• Refactored optimization record-handling and chart-saving logic.

• Updated schema dependencies for backward compatibility with legacy optimization flags.

• Simplify optimization logic, add optimization flag, and update default method in DftbInput (e1846f1)

• Removed NONE from OptimizationMethod enum.

• Added optimization boolean field to indicate geometry optimization.

• Updated default optimization method to steepest_descendent.

• Refactored schema and UI dependencies to streamline configuration.

• Updated Dockerfile with symbolic links for legacy Simstack compatibility.

• Adjusted Python version requirement in uv.lock and updated Simstack dependency.

---

Detailed Changes: v0.6.0...v0.7.0

v0.6.0

v0.6.0 (2026-08-24)

This release is published under the MIT License.

Features

• Build DFTB+ from source with Python API support and update Simstack dependency (0968df1)

---

Detailed Changes: v0.5.1...v0.6.0

v0.5.1

v0.5.1 (2026-08-24)

This release is published under the MIT License.

Bug Fixes

• Isolate DFTB+ jobs so concurrent runs cannot abort the container (e16d8a9)

INSTANCE_SAFE_BUILD does not namespace Fortran files. Spawn each calculation in its own working directory and surface dftbplus.log when Fortran error stop kills the worker.

Co-authored-by: Cursor cursoragent@cursor.com

---

Detailed Changes: v0.5.0...v0.5.1