# NousResearch/hermes-agent · GPT-6 Astra reasoning effort clamping on Codex

## Open now

- Nothing open on GitHub.

These items concern GPT-6 Astra's availability, opt-in context limits, and a bug where MAX reasoning effort is silently downgraded to XHIGH when using the openai-codex backend. A fix to preserve the intended effort level is included.

## TL;DR

- 3 member(s): 0 open issue(s), 1 closed issue(s), 0 open PR(s), 1 merged PR(s), 1 PR(s) closed without merging.
- A fix was merged to main in PR #105185. Verify it addresses your case.
- **Recommended:** read PR #105185 (and any competing PRs below) before filing. Confirm the merged change covers your symptom.

## Maintainer resolution verdict

- Internal-tracker verdict over this cluster's issue members: `unresolved`. This is distinct from the fix status above. Weigh it yourself.

## How this cluster was grouped

- Link evidence (`STRUCTURAL_SPANNING`): every member is connected by structural evidence: shared changed files, or a closing/superseded reference.
- Pairwise link density `0.67`: the share of member pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36. `1.00` means every pair is linked.

## Closed / duplicate issues

- #103556: \[Bug\]: gpt-6-astra MAX reasoning is silently clamped to XHIGH on openai-codex

## Prior & competing PRs

- **#105185: GPT-6 Astra: discovery-gated availability, Codex 272K/900K opt-in, effort + pricing contract (salvage #103057)**. merged to main · fix tier `verified`
  - maintainer direction: category `merged`
  - files: `agent/auxiliary_client.py`, `agent/model_metadata.py`, `agent/models_dev.py`, `agent/reasoning_effort.py`, `agent/transports/codex.py`, `agent/usage_pricing.py`, `hermes_cli/codex_models.py`, `hermes_cli/inventory.py`, `hermes_cli/model_switch_providers.py`, `hermes_cli/models.py`, `tests/agent/test_astra_baseline_runtime.py`, `tests/agent/test_auxiliary_client.py`, `tests/agent/test_models_dev.py`, `tests/agent/test_usage_pricing.py`, `tests/agent/transports/test_codex_transport.py`, `tests/hermes_cli/test_codex_models.py`, `tests/hermes_cli/test_openai_picker_curated.py`
- **#103570: fix(codex): preserve max reasoning effort for gpt-6-astra on openai-codex (#103556)**. closed without merge
  - maintainer direction: category `implemented_elsewhere`; implication `superseded`; refs `233757037df1f03f9fe1cfddc097acd5ad7f7510`, `#105185`

## Fix status

- Merged to main in PR #105185 (cited above). No release data is tracked. Verify downstream.
- Fix last verified (merged) `2026-09-07`.