Release v1.0.1

Fixes:
- CI/CD workflows (MSRV 1.85.0, codeql-action, TruffleHog)
- Code quality (clippy warnings, formatting)

All GitHub Actions workflows now passing.

v1.0.0: Production Release

Infrastructure-first release with comprehensive RSR compliance, documentation,
tests, and CI/CD. Production-ready for X-Ray analysis and stress testing.

Key Features:
- Zero duplicate weak points (per-file analysis)
- All locations populated (never null)
- Per-file breakdown with risk scoring
- Latin-1 fallback for non-UTF-8 files
- Pattern library and RuleSet fully wired
- 21 tests passing, zero warnings

Infrastructure:
- RSR compliant (AI manifest, SCM files, workflows)
- 11 GitHub Actions workflows
- Comprehensive documentation (SECURITY, CONTRIBUTING, LICENSE)
- Stable JSON schema v1.0
- Code coverage reporting
- MSRV 1.75.0

Quality Metrics:
- 3,200+ lines of Rust
- 21/21 tests passing
- 0 compiler warnings
- OpenSSF Scorecard ready
- CodeQL analysis enabled

Tested Against:
- echidna: 15 weak points (down from 271 in v0.1)
- eclexia: 7 weak points
- panic-attacker itself: ≤5 weak points

Future work (v1.x/v2.0):
- Constraint sets (YAML stress profiles)
- Real Datalog engine (Crepe/Datafrog)
- Multi-program testing
- Language expansion

This release prioritizes solid foundation over feature quantity.

v0.2.0: Quality fixes and per-file analysis

Major improvements to X-Ray analysis quality:
- Per-file statistics eliminate duplicates (271→15 weak points on echidna)
- All weak points include file locations (never null)
- Latin-1 fallback for non-UTF-8 source files
- Verbose mode with per-file breakdown sorted by risk
- Pattern library and RuleSet wired in
- Zero compiler warnings
- 7/7 tests passing (3 new integration tests)

This release makes the output trustworthy and ready for production use.