fp-tools v0.2.7

fp-tools v0.2.7 improves workflow setup, progress reporting, and comparison validation.

• bulk-footprinting --genome hg38 and --genome mm10 use the matching managed reference and blacklist, with visible download, checksum, cache, and indexing progress. Custom references still require an explicit blacklist.
• Omitted core limits use all available processor cores. Workflow wrappers and graphical forms preserve automatic selection in saved YAML; an explicit limit remains available.
• Bulk, single-cell, preprocessing, and YAML workflows display live stage diagnostics while retaining their log files.
• Fixed #76: duplicate or reversed condition pairs fail before reference setup or analysis when a combined review is requested. Errors identify comparison IDs and table lines. Use --review-format none for separate reports with repeated condition labels.
• Invalid combined-review generation preserves existing report output.
• Simplified documentation examples and removed redundant CI steps while retaining platform and release-artifact checks.

Production scientific methods and defaults are unchanged. Research-branch methods are not included.

Install the Python package with python -m pip install --upgrade fp-tools-bio==0.2.7, or use the Windows executable or Apple Silicon DMG below. Check downloaded desktop files against SHA256SUMS.txt; managed runtimes and container archives include their own checksums.

Validation completed:

• Cross-platform CI: all ten jobs passed. Linux: 604 tests and 87 subtests passed; Windows Python 3.12: 592 tests and 69 subtests passed, with Python 3.11 and 3.13 also passing.
• Native desktop validation: Windows and Apple Silicon GUI, command dispatch, reports, and packaging checks passed. An audit-only dropdown fix was applied after tagging; packaged application source is identical to the tag.
• Public-download discovery: both Windows cases and macOS passed fresh-cache and cached execution.
• Managed runtimes, containers, and PyPI publication passed.
• All 13 GitHub binary/archive downloads and all 16 PyPI files matched their checksums. A clean PyPI installation passed 76 focused tests and a real five-stage small bulk workflow, numerical comparison, and resume check.
• Strict documentation builds, browser audits, console checks, and 18 example YAML dry runs passed.

These checks include native desktop execution and small scientific fixtures; they do not certify full-scale single-cell analyses.