Packages

codex_sdk

0.2.2
0.17.0 0.16.1 0.16.0 0.14.0 0.13.0 0.12.0 0.11.0 0.10.1 0.10.0 0.9.0 0.8.0 0.7.2 0.7.1 0.7.0 0.6.0 0.5.0 0.4.5 0.4.4 0.4.3 0.4.2 0.4.1 0.4.0 0.3.0 0.2.5 0.2.4 0.2.3 0.2.2 0.2.1 0.2.0 0.1.0

Idiomatic Elixir SDK for OpenAI's Codex agent. Provides a complete, production-ready interface with streaming support, comprehensive event handling, and robust testing utilities.

HexDocs

Current section

Files

Jump to
Readme
30 Versions
6 Dependencies
Dependants
Files
Activity
codex_sdk docs 20251129 models models_compaction_and_prompts.md
Raw

docs/20251129/models/models_compaction_and_prompts.md 

# Models, compaction, and prompt handling changes for the Elixir SDK



Upstream changes:

- New model defaults and reasoning-level tweaks (gpt-5.1 family) shipped; some experimental tool-enabled models were added.

- Remote compaction is on by default with explicit compaction events and token-usage updates emitted during turns.

- Truncation helpers were refactored to avoid double-truncation and to tokenize more accurately during unified exec and history replay.



SDK impact:

- Validate that default model selection and reasoning options in the SDK match the upstream CLI defaults.

- Compaction now surfaces explicit events and token deltas; SDK consumers should be ready to handle those streams.

- Any SDK-side truncation or history replay should align with the refactored helpers to avoid mismatched token accounting.



Action items:

- Refresh the model list/constants exposed by the SDK and document reasoning effort defaults.

- Add support for compaction events and token-usage updates in event handling and tests.

- Revisit truncation logic (or delegate to the CLI) to mirror upstream behavior and prevent drift in prompt sizing.