fix: suppress verbose output in tests

- Fix MD060 table formatting in CHANGELOG.md
- Replace @info logs with silent comments for CUDA GPU skips
- Redirect stdout for show_time=true test to suppress timing output
- Keep test logic intact while reducing noise

Author: ocots

BREAKING.md

@@ -5,13 +5,34 @@ and provides migration guides for users upgrading between versions.
 
 ---
 
-## v0.3.1-beta (2026-02-14)
+## v0.3.3-beta (2026-02-16)
+
+**Breaking change:** The base solver abstract type was renamed from
+`AbstractOptimizationSolver` to `AbstractNLPSolver` for consistency with the
+`AbstractNLPModeler` naming introduced in v0.3.0.
+
+### Migration
+
+Replace any references to the old abstract type:
 
-**No breaking changes** — This release only adds enhancements:
+```text
+AbstractOptimizationSolver → AbstractNLPSolver
+```
 
-- Backend override options now accept both `Type{<:ADBackend}` and `ADBackend` instances
-- Improved validation and error messages
-- Better test coverage and documentation
+No other API changes are required.
+
+---
+
+## v0.3.2-beta (2026-02-15)
+
+No breaking changes. This release focused on options getters/encapsulation
+and documentation updates.
+
+---
+
+## v0.3.1-beta (2026-02-14)
+
+No breaking changes.
 
 ---
 
@@ -33,25 +54,25 @@ This aligns with Julia conventions (`Module.Type`) and improves readability.
 
 ### Modelers
 
-| v0.2.x | v0.3.0 |
-|--------|--------|
-| `ADNLPModeler` | `Modelers.ADNLP` |
-| `ExaModeler` | `Modelers.Exa` |
-| `AbstractOptimizationModeler` | `AbstractNLPModeler` |
+| v0.2.x                       | v0.3.0                 |
+|------------------------------|------------------------|
+| `ADNLPModeler`               | `Modelers.ADNLP`       |
+| `ExaModeler`                 | `Modelers.Exa`         |
+| `AbstractOptimizationModeler`| `AbstractNLPModeler`   |
 
 ### Solvers
 
-| v0.2.x | v0.3.0 |
-|--------|--------|
-| `IpoptSolver` | `Solvers.Ipopt` |
-| `MadNLPSolver` | `Solvers.MadNLP` |
-| `MadNCLSolver` | `Solvers.MadNCL` |
-| `KnitroSolver` | `Solvers.Knitro` |
+| v0.2.x        | v0.3.0           |
+|---------------|------------------|
+| `IpoptSolver` | `Solvers.Ipopt`  |
+| `MadNLPSolver`| `Solvers.MadNLP` |
+| `MadNCLSolver`| `Solvers.MadNCL` |
+| `KnitroSolver`| `Solvers.Knitro` |
 
 ### DOCP
 
-| v0.2.x | v0.3.0 |
-|--------|--------|
+| v0.2.x                             | v0.3.0             |
+|------------------------------------|--------------------|
 | `DiscretizedOptimalControlProblem` | `DiscretizedModel` |
 
 ---
@@ -62,7 +83,7 @@ This aligns with Julia conventions (`Module.Type`) and improves readability.
 
 The simplest migration is a global search-and-replace in your codebase:
 
-```
+```text
 ADNLPModeler                      →  Modelers.ADNLP
 ExaModeler                        →  Modelers.Exa
 AbstractOptimizationModeler       →  AbstractNLPModeler

CHANGELOG.md

@@ -1,4 +1,5 @@
 # Changelog
+<!-- markdownlint-disable MD024 -->
 
 All notable changes to CTSolvers.jl will be documented in this file.
 
@@ -9,6 +10,41 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.3.3-beta] - 2026-02-16
+
+### Changed
+
+- **Solver abstract type rename** — `AbstractOptimizationSolver` was renamed to
+  `AbstractNLPSolver` for consistency with `AbstractNLPModeler` naming
+- **Docs maintenance** — Updated references to the new abstract solver type
+  across orchestration/routing examples and solver documentation
+
+### Fixed
+
+- **Test alignment** — Tests updated to use `AbstractNLPSolver`, keeping
+  inheritance and contract checks consistent with the new naming
+
+---
+
+## [0.3.2-beta] - 2026-02-15
+
+### Added
+
+- **Options getters** — New getters/exported helpers for `StrategyOptions`
+
+### Changed
+
+- **Encapsulation** — Internal access to strategy options now goes through
+  `_raw_options`/getter helpers; docs updated accordingly
+- **Docs** — Options system guide expanded and translated to English sections
+
+### Fixed
+
+- **Test refactor** — Tests updated to use the new getters and encapsulation
+  pattern
+
+---
+
 ## [0.3.1-beta] - 2026-02-14
 
 ### Added
@@ -25,6 +61,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   - `jacobian_backend`, `hessian_backend`, `ghjvprod_backend`
   - From: `Union{Nothing, ADNLPModels.ADBackend}`
   - To: `Union{Nothing, Type{<:ADNLPModels.ADBackend}, ADNLPModels.ADBackend}`
+- **Solver abstract type rename** — `AbstractOptimizationSolver` was renamed to
+  `AbstractNLPSolver` for consistency with `AbstractNLPModeler` naming
 - **Validation logic** — `validate_backend_override()` now correctly handles three forms:
   - `nothing` (use default)
   - `Type{<:ADBackend}` (constructed by ADNLPModels)
@@ -41,6 +79,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Technical Details
 
 #### Backend Override Usage
+
 ```julia
 # Three accepted forms:
 Modelers.ADNLP(gradient_backend=nothing)                              # Use default
@@ -49,6 +88,7 @@ Modelers.ADNLP(gradient_backend=ADNLPModels.ForwardDiffADGradient())  # Instance
 ```
 
 #### Type Declaration Change
+
 ```julia
 # Before
 type=Union{Nothing, ADNLPModels.ADBackend}
@@ -124,6 +164,7 @@ See [BREAKING.md](BREAKING.md) for a detailed migration guide.
 ### Technical Details
 
 #### Import Refactoring
+
 ```julia
 # Before
 using DocStringExtensions
@@ -135,6 +176,7 @@ import NLPModels
 ```
 
 #### GPU Test Implementation
+
 ```julia
 # Before (dead code)
 # using MadNLPGPU
@@ -146,6 +188,7 @@ gpu_solver = Solvers.MadNLP(linear_solver=MadNLPGPU.CUDSSSolver)
 ```
 
 #### CUDA Availability Helper
+
 ```julia
 is_cuda_on() = CUDA.functional()
 if is_cuda_on()
@@ -205,6 +248,7 @@ end
 ### 🎉 BREAKING CHANGES
 
 ### Added
+
 - **New option validation system** with strict and permissive modes
 - **`mode::Symbol` parameter** to strategy constructors (`:strict` default, `:permissive`)
 - **`route_to()` helper function** for option disambiguation
@@ -213,34 +257,40 @@ end
 - **Comprehensive test suite** with 66 tests covering all scenarios
 
 ### Changed
+
 - **`build_strategy_options()`** now supports `mode` parameter
 - **`route_all_options()`** now supports `mode` parameter
 - **Error handling** uses CTBase `Exceptions.IncorrectArgument` and `Exceptions.PreconditionError`
 - **Warning system** for unknown options in permissive mode
 - **Documentation** completely updated with examples and tutorials
 
 ### Deprecated
+
 - **Tuple syntax for disambiguation** (still supported but deprecated)
   - Old: `max_iter = (1000, :solver)`
   - New: `max_iter = route_to(solver=1000)`
 
 ### Fixed
+
 - **Option validation** now provides helpful error messages
 - **Disambiguation** works clearly with `route_to()`
 - **Type safety** improved with `RoutedOption` type
 - **Memory usage** optimized for validation system
 
 ### Security
+
 - **Strict mode by default** prevents unknown option errors
 - **Input validation** enhanced with type checking
 - **Error messages** don't leak sensitive information
 
 ### Performance
+
 - **Minimal overhead**: < 1% for strict mode, < 5% for permissive mode
 - **Type stability** maintained throughout validation system
 - **Memory efficiency** optimized for large option sets
 
 ### Documentation
+
 - **Complete user guide** with examples and best practices
 - **Migration guide** for existing code
 - **API reference** with detailed examples
@@ -252,13 +302,15 @@ end
 ## [0.1.0] - 2025-XX-XX
 
 ### Added
+
 - Initial release of CTSolvers.jl
 - Basic strategy construction and management
 - Option handling and validation
 - Strategy registry and metadata system
 - Integration with NLPModels and solvers
 
 ### Features
+
 - Strategy builders and constructors
 - Option extraction and validation
 - Strategy registry with metadata
@@ -354,7 +406,7 @@ use permissive mode:
 ## Performance Impact
 
 | Operation | Before | After (Strict) | After (Permissive) | Overhead |
-|-----------|--------|----------------|-------------------|----------|
+| ----------- | -------- | ---------------- | -------------------- | ---------- |
 | Strategy construction | 100μs | 101μs | 105μs | < 1% / < 5% |
 | Option validation | 50μs | 50μs | 52μs | 0% / < 4% |
 | Disambiguation | N/A | 1μs | 1μs | < 1% |
@@ -390,6 +442,6 @@ use permissive mode:
 
 ## Questions?
 
-- **GitHub Issues**: https://github.com/control-toolbox/CTSolvers.jl/issues
-- **Discord**: https://discord.gg/control-toolbox
-- **Documentation**: https://control-toolbox.github.io/CTSolvers.jl/
+- **GitHub Issues**: <https://github.com/control-toolbox/CTSolvers.jl/issues>
+- **Discord**: <https://discord.gg/control-toolbox>
+- **Documentation**: <https://control-toolbox.github.io/CTSolvers.jl/>

test/suite/extensions/test_madncl_extension.jl

@@ -497,7 +497,7 @@ function test_madncl_extension()
                 #     Test.@test Array(sol.solution)[1] ≈ max_prob.sol[1] atol=1e-6
                 # end
             else
-                @info "CUDA not functional, skipping GPU tests."
+                # CUDA not functional — skip silently (reported in runtests.jl)
             end
         end
 
@@ -537,7 +537,7 @@ function test_madncl_extension()
                 #     Test.@test Array(sol.solution)[1] ≈ max_prob.sol[1] atol=1e-6
                 # end
             else
-                @info "CUDA not functional, skipping GPU solve_with_madncl tests."
+                # CUDA not functional — skip silently (reported in runtests.jl)
             end
         end
 
@@ -570,7 +570,7 @@ function test_madncl_extension()
                     Test.@test Array(sol.solution) ≈ expected atol=1e-6
                 end
             else
-                @info "CUDA not functional, skipping GPU initial guess tests."
+                # CUDA not functional — skip silently (reported in runtests.jl)
             end
         end
     end

test/suite/optimization/test_optimization.jl

@@ -142,8 +142,10 @@ function test_optimization()
                 Test.@test nlp.meta.x0 == x0
 
                 # Test with kwargs
-                nlp2 = builder(x0; show_time=true)
-                Test.@test calls[] == 2
+                redirect_stdout(devnull) do
+                    nlp2 = builder(x0; show_time=true)
+                    Test.@test calls[] == 2
+                end
             end
 
             Test.@testset "ExaModelBuilder" begin