Support ModelManager's post_processor hook; UQ docs

[drbergman]

Summary

• ModelManager now runs a user post_processor callback between two simulator hooks: a non-destructive postSimulationProcessing and a new destructive postSimulationCleanup, so a callback always sees intact simulation output before PCMM prunes it. This moves PCMM's pruning (previously in postSimulationProcessing) into the new postSimulationCleanup.
• Fixes a latent bug found while testing this: pathToOutputFolder(::Integer) in src/loader.jl shadowed ModelManager's re-exported generic function instead of extending it, making the SimulationProcess-argument method (needed by a post_processor) unreachable from PCMM's namespace. Removed — PCMM now relies purely on ModelManager's pathToOutputFolder.
• Docs: adds a monadsTable section (new ModelManager export), documents batch run(Vector) and the max_evaluations budget behavior (now enforced before each batch, and counts particles/monads not simulations), and renames the docs nav section "Experiments" → "Uncertainty Quantification" to match ModelManager's naming.
• Bumps PhysiCellModelManager to 0.3.3.

⚠️ Dependency note

This branch was developed and tested against a local dev checkout of ModelManager (the version with postSimulationCleanup / the reordered hooks / monadsTable is not yet registered). CI on this PR will likely fail to resolve/precompile against the currently registered ModelManager 0.7.x until that version is released. Do not treat that as a regression in this PR — hold merge until the corresponding ModelManager release lands, then re-run CI.

Test plan

• Full local suite run (Pkg.test()) against the ModelManager dev checkout, including a new PrunerTests.jl case: a post_processor reading pathToOutputFolder(sp) sees output present (via output*.mat files) during the callback, and confirms it's pruned only afterward.
• Existing no-callback pruning assertions (Loader/Population/Substrate tests) still pass — no regression when post_processor isn't used.
• Re-run CI once a ModelManager release with the reordered hooks is available.

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR aligns PhysiCellModelManager.jl with ModelManager’s new per-simulation hook ordering by moving PCMM’s destructive “pruning + err-file handling” work into postSimulationCleanup, ensuring a user post_processor callback runs against intact simulation outputs. It also removes a pathToOutputFolder(::Integer) definition that previously shadowed (rather than extended) ModelManager’s generic, and updates UQ documentation (including monadsTable, batch run(Vector), and max_evaluations semantics) while bumping the package version.

Changes:

• Move destructive post-simulation behavior from postSimulationProcessing to postSimulationCleanup so post_processor sees un-pruned outputs.
• Remove pathToOutputFolder(::Integer) from src/loader.jl to avoid shadowing ModelManager’s API.
• Add/adjust docs for monadsTable, batched run(Vector), and max_evaluations behavior; bump version to 0.3.3.

Reviewed changes

Copilot reviewed 10 out of 11 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
test/test-scripts/PrunerTests.jl	Adds a regression test asserting outputs are intact during post_processor and pruned after cleanup.
src/simulator_interface.jl	Switches PCMM’s hook implementation to extend postSimulationCleanup and updates hook docs.
src/loader.jl	Removes a local pathToOutputFolder(::Integer) that shadowed ModelManager’s generic.
README.md	Documents the new post_processor hook behavior and where results are stored/read.
Project.toml	Bumps package version to 0.3.3.
progress.md	Adds a development log entry describing the hook split and doc updates.
PRD.md	Adds a feature spec section describing the post-processing hook contract and QoI behavior.
docs/src/man/querying_parameters.md	Documents monadsTable as the monad-level analogue of simulationsTable.
docs/src/man/examples.md	Adds recipe for batching pre-built trials into a single run(Vector) call.
docs/src/man/calibration.md	Clarifies max_evaluations enforcement and that it counts particles/monads, not simulations.
docs/make.jl	Renames nav section “Experiments” → “Uncertainty Quantification”.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!