Bump version to 3.8.0

[igerber]

Summary

• Bump version 3.7.0 → 3.8.0 across all surfaces (__init__.py, pyproject.toml, rust/Cargo.toml, llms-full.txt, CITATION.cff).
• Promote the curated [Unreleased] CHANGELOG section to ## [3.8.0] - 2026-07-18 and add the v3.7.0...v3.8.0 comparison link.

What this release ships (already merged to main)

• RegressionDiscontinuity (alias RDD) — sharp, fuzzy, and covariate-adjusted local-polynomial regression discontinuity with robust bias-corrected inference, parity-targeting R rdrobust 4.0.0 end-to-end (all 10 bandwidth selectors, triangular/epanechnikov/uniform kernels, mass-points, fuzzy Wald ratio via treatment_col=, additive covariate adjustment via covariates=, with guarded degenerate-design handling documented as deviations from R). Merged in docs: add RDD paper reviews (CCT 2014, CCFT 2017, CCF 2018, CCFT 2019) #675/feat: rdrobust 4.0.0 sharp-RD bandwidth-selection port + R golden parity (RDD PR-1) #680/feat: RegressionDiscontinuity - sharp RD estimator with rdrobust 4.0.0 parity (RDD PR-2) #684/feat: fuzzy regression discontinuity via treatment_col on RegressionDiscontinuity #686/feat(rdd): covariate-adjusted RD - covariates= on fit() with rdrobust 4.0.0 parity (CCFT 2019) #691.
• Internal: mypy enforced at zero errors across the package (chore: pin lint tooling + normalize source and hand-fixed test files (1/2) #677/chore: mechanical black/ruff normalization of remaining test files (2/2) #679/ci: ungated check-only Lint workflow (ruff + black) + blame-ignore + docs #681/chore(types): mypy triage to enforceable zero + blocking CI job (3/3) #683).

Methodology references

• N/A — this PR contains no estimator/math changes. It is a release version bump only (version strings + CHANGELOG/CITATION). The RDD methodology and its citations (Calonico, Cattaneo & Titiunik 2014; CCFT 2019) already landed in the feature PRs above and are documented in docs/methodology/REGISTRY.md.

Validation

• Tests added/updated: none — release bump only. The RDD test suites (tests/test_rdd*.py, tests/test_rdrobust_port.py) and the mypy-zero gate are already on main and green.
• black --check and ruff check pass on the touched Python file (diff_diff/__init__.py).
• Version consistency verified across all six surfaces.

Security / privacy

• Confirm no secrets/PII in this PR: Yes

---

🤖 Generated with Claude Code

[github-actions]

PR Review Report

Overall Assessment

✅ Looks good

No unmitigated P0/P1 findings.

Executive Summary

• The diff is limited to release metadata, changelog promotion, and LLM guide version text.
• No estimator, math, weighting, variance/SE, identification assumption, or default behavior changed.
• RegressionDiscontinuity/RDD release-note claims are backed by docs/methodology/REGISTRY.md and diff_diff/rdd.py docstrings.
• Version surfaces are synchronized across Python, Rust, CFF, changelog, and guide metadata.
• One minor documentation-maintenance note below; not blocking.

Methodology

No findings.

• Severity: None
• Impact: No affected methods. The PR does not modify estimator implementation or inference behavior. The RDD methodology referenced in CHANGELOG.md:L8-L24 is documented in docs/methodology/REGISTRY.md and in-code docstrings.
• Concrete fix: None.

Code Quality

No findings.

• Severity: None
• Impact: Metadata-only changes are straightforward and internally consistent: diff_diff/__init__.py:L319, pyproject.toml:L7, rust/Cargo.toml:L3, CITATION.cff:L10-L11, diff_diff/guides/llms-full.txt:L5.
• Concrete fix: None.

Performance

No findings.

• Severity: None
• Impact: No runtime or performance-sensitive code changed.
• Concrete fix: None.

Maintainability

No findings.

• Severity: None
• Impact: No new branching, duplication, API surface, or generated-code dependency introduced.
• Concrete fix: None.

Tech Debt

No findings.

• Severity: None
• Impact: No new deferred correctness issue or TODO-dependent mitigation is introduced.
• Concrete fix: None.

Security

No findings.

• Severity: None
• Impact: No credentials, network behavior, parsing surface, or executable logic changed.
• Concrete fix: None.

Documentation/Tests

• Severity: P3 informational
• Impact: CHANGELOG.md:L8 replaces [Unreleased] with [3.8.0], leaving no empty [Unreleased] section for future entries. This is only a changelog-maintenance issue and does not affect release correctness.
• Concrete fix: Optionally add an empty ## [Unreleased] section above ## [3.8.0] - 2026-07-18 and add an [Unreleased] compare link from v3.8.0...HEAD.

No tests are required for this metadata-only PR.