Add body_after_prefix auto-compact token limit scope (#22870)

## Why

`model_auto_compact_token_limit` has only been able to budget the full
active context. That makes it hard to set a small "growth since
compaction" budget for sessions that preserve a large carried window
prefix: the preserved prefix can consume the whole budget and force
immediate repeated compaction.

This PR adds an opt-in `body_after_prefix` scope so callers can apply
`model_auto_compact_token_limit` to sampled output and later growth
after the current carried prefix, while still forcing compaction before
the full model context window is exhausted.

## What changed

- Adds `AutoCompactTokenLimitScope` with the existing `total` behavior
as the default and a new `body_after_prefix` mode:
[`config_types.rs`](https://github.com/openai/codex/blob/973806b1cb35792555bead994cb3ed94656eb171/codex-rs/protocol/src/config_types.rs#L24-L37).
- Threads `model_auto_compact_token_limit_scope` through config loading,
`Config`, `core-api`, and app-server v2 schema/TypeScript generation.
- Records the first observed input-token count for a `body_after_prefix`
compaction window and uses it as the baseline when deciding whether the
scoped auto-compaction budget is exhausted:
[`turn.rs`](https://github.com/openai/codex/blob/973806b1cb35792555bead994cb3ed94656eb171/codex-rs/core/src/session/turn.rs#L743-L781).
- Keeps a hard context-window cap in `body_after_prefix`, so scoped
budgeting cannot let the active context overrun the usable window.

## Verification

Added compact-suite coverage for the two key behaviors:
`body_after_prefix` does not re-compact just because the carried prefix
is larger than the scoped budget, and it still compacts when the total
active context reaches the configured context window:
[`compact.rs`](https://github.com/openai/codex/blob/973806b1cb35792555bead994cb3ed94656eb171/codex-rs/core/tests/suite/compact.rs#L3003-L3128).

codex-rs/core/src/config/config_tests.rs

@@ -7711,6 +7711,7 @@ async fn test_precedence_fixture_with_o3_profile() -> std::io::Result<()> {
             review_model: None,
             model_context_window: None,
             model_auto_compact_token_limit: None,
+            model_auto_compact_token_limit_scope: AutoCompactTokenLimitScope::Total,
             service_tier: None,
             model_provider_id: "openai".to_string(),
             model_provider: fixture.openai_provider.clone(),
@@ -8164,6 +8165,7 @@ async fn test_precedence_fixture_with_gpt3_profile() -> std::io::Result<()> {
         review_model: None,
         model_context_window: None,
         model_auto_compact_token_limit: None,
+        model_auto_compact_token_limit_scope: AutoCompactTokenLimitScope::Total,
         service_tier: None,
         model_provider_id: "openai-custom".to_string(),
         model_provider: fixture.openai_custom_provider.clone(),
@@ -8331,6 +8333,7 @@ async fn test_precedence_fixture_with_zdr_profile() -> std::io::Result<()> {
         review_model: None,
         model_context_window: None,
         model_auto_compact_token_limit: None,
+        model_auto_compact_token_limit_scope: AutoCompactTokenLimitScope::Total,
         service_tier: None,
         model_provider_id: "openai".to_string(),
         model_provider: fixture.openai_provider.clone(),
@@ -8483,6 +8486,7 @@ async fn test_precedence_fixture_with_gpt5_profile() -> std::io::Result<()> {
         review_model: None,
         model_context_window: None,
         model_auto_compact_token_limit: None,
+        model_auto_compact_token_limit_scope: AutoCompactTokenLimitScope::Total,
         service_tier: None,
         model_provider_id: "openai".to_string(),
         model_provider: fixture.openai_provider.clone(),

codex-rs/core/src/config/mod.rs

@@ -77,6 +77,7 @@ use codex_model_provider_info::built_in_model_providers;
 use codex_model_provider_info::merge_configured_model_providers;
 use codex_models_manager::ModelsManagerConfig;
 use codex_protocol::config_types::AltScreenMode;
+use codex_protocol::config_types::AutoCompactTokenLimitScope;
 use codex_protocol::config_types::ForcedLoginMethod;
 use codex_protocol::config_types::Personality;
 use codex_protocol::config_types::ReasoningSummary;
@@ -527,6 +528,10 @@ pub struct Config {
     /// Token usage threshold triggering auto-compaction of conversation history.
     pub model_auto_compact_token_limit: Option<i64>,
 
+    /// Controls whether `model_auto_compact_token_limit` applies to the full
+    /// active context or only tokens after the carried compaction-window prefix.
+    pub model_auto_compact_token_limit_scope: AutoCompactTokenLimitScope,
+
     /// Key into the model_providers map that specifies which provider to use.
     pub model_provider_id: String,
 
@@ -3373,6 +3378,9 @@ impl Config {
             review_model,
             model_context_window: cfg.model_context_window,
             model_auto_compact_token_limit: cfg.model_auto_compact_token_limit,
+            model_auto_compact_token_limit_scope: cfg
+                .model_auto_compact_token_limit_scope
+                .unwrap_or_default(),
             model_provider_id,
             model_provider,
             cwd: resolved_cwd,