v0.3.0 - Resilient and Observable RLM Runs

Highlights

• Added budget-aware retries with exact call and cost accounting.
• Added final-answer validation with correction feedback.
• Added typed success and failure results with JSONL export.
• Preserved partial trajectories and usage statistics for failed runs.
• Improved handling of null and malformed provider responses.
• Added reproducible 1M-character and multi-document benchmarks.

Validation

• 178 tests passing
• 93.5% coverage with branch tracking
• Verified across Python 3.9 to 3.12 on Linux, macOS, and Windows

See CHANGELOG.md and BENCHMARK_RESULTS.md for details.

v0.2.0 - Bounded and Observable RLM Runs

Highlights

• Added tree-wide call, token, cost, and elapsed-time budgets.
• Added structured results with per-run statistics and execution trajectories.
• Aligned recursion-depth semantics with the RLM paper.
• Hardened REPL execution and added optional resource limits.
• Added reproducible long-context benchmarks and direct-model baselines.
• Added CI for Python 3.9-3.12 across Linux, macOS, and Windows.
• Expanded the test suite to 135 tests with enforced branch coverage.

Compatibility

• Fixed REPL comprehension behavior on Python 3.9-3.11.
• Verified the complete test suite on all supported platforms.
• Updated GitHub Actions to Node 24-based releases.

See CHANGELOG.md and BENCHMARK_RESULTS.md for details.

v0.1.0 - Initial Release

Initial implementation of Recursive Language Models

Features:

• Process 100k+ token contexts
• Support for 100+ LLM providers via LiteLLM
• Async and sync APIs
• Safe REPL execution with RestrictedPython
• Two-model optimization support
• Comprehensive test suite (43 tests, 100% passing)