Codex installs now match the model you actually run.
- gpt-5.6-sol model gets a bounded-scope profile that finishes the explicit task and stops, treating adjacent findings as report-only
- setup command reads the top-level model from ${CODEX_HOME:-~/.codex}/config.toml and renders the matching profile when called with --host codex
- Resolver warns when a near-miss model ID like gpt-5.6-sol-2026-08-01 falls back to generic gpt profile
- Generated upgrade skills reinstall their own host via ./setup --host codex in Codex and --host kiro in Kiro
- Kiro installs render the claude profile before copying skills, then restore the resolved Codex profile to prevent GPT-family behavioral text from shipping
- Codex skills path honors $CODEX_HOME environment variable
- Hermetic Codex E2E runner copies auth.json only, preventing operator plugins, MCP servers, rules, and skills from leaking into evals
- Hermetic Codex E2E runner supports per-run model, TOML config overrides, and --ignore-user-config flag
- Every gstack skill behavioral patch is now model-aware for Codex instead of generic
- setup command resolves the Codex generation model on every run as a read-only TOML lookup to preserve a Sol user's rendered profile
- Codex install summary now prints the active profile and its source
From gstack
gpt-5.6-sol gets a bounded-scope profile that finishes the job, then stops.
Every gstack skill carries a model-specific behavioral patch. This release makes that patch model-aware for Codex: ./setup --host codex reads the top-level model from ${CODEX_HOME:-~/.codex}/config.toml and renders the matching profile.
The headline is gpt-5.6-sol. Sol reads completeness language like "exhaustive" and "Boil the Ocean" as authority to keep going, widening into adjacent cleanup and speculative hardening nobody asked for. Its new profile pins the boundary: the explicit task is the lake, adjacent findings are report-only, investigation stops once the cause is established, and the run terminates on one clean verification pass.
Full coverage inside the boundary still applies, and the AskUserQuestion decision-brief format is never trimmed. Sol is exact-match only. Terra, Luna, dated snapshots, and any suffixed ID deliberately fall back to the generic GPT profile, and the resolver warns when a near-miss like gpt-5.6-sol-2026-08-01 lands on generic gpt.
Generated upgrade skills reinstall their own host: ./setup --host codex in Codex renders, --host kiro in Kiro copies (rewritten at copy time), bare ./setup only for Claude.
Kiro installs render the claude profile before copying skills, then restore the resolved Codex profile, so Kiro never ships GPT-family behavioral text and live ~/.codex symlinks stay correct. The Codex skills path honors $CODEX_HOME.
The hermetic Codex E2E runner copies auth.json only. Operator plugins, MCP servers, rules, and skills no longer leak into supposedly hermetic evals. Per-run model, TOML config overrides, and --ignore-user-config are supported.
setup resolves the Codex generation model on every run (a read-only TOML lookup), so any install path preserves a Sol user's rendered profile; the codex install summary prints the active profile and its source.