Retire the personality feature flag and document deprecated settings (#45809)

## What changed

- Accept and ignore `features.personality` in user configuration, profiles, and managed requirements.
- Remove feature gating from `personality = "none"`, which strips the literal `# Personality` section when preparing model catalog instructions. Preserve explicit base instructions and existing thread instructions.
- Document deprecated personality fields in configuration, app-server schemas, and the Python SDK: `friendly` and `pragmatic` no longer select a style, and `supportsPersonality` is always `false`.

## Testing

Add regression coverage for ignored legacy flag values and managed requirements, personality opt-out behavior with the flag absent or set to either value, role overrides, and preservation of explicit base instructions, including empty strings.

GitOrigin-RevId: 4e12c66b42bfc59d6f151a5b3c28fadc6654ae99
This commit is contained in:
rhan-oai authored and copyberry committed 2026-09-15 23:39:40 +00:00
1 parent 4d2807023a
commit 883af106b9
54 files changed
+389 -264

No files matched your search

@@ -11,7 +11,7 @@ fn mock_responses_config_composes_model_provider_features_and_extra_tables() {
.with_provider_base_url("http://127.0.0.1:1234/api/codex")
.with_approval_policy("on-request")
.with_sandbox_mode("workspace-write")
.enable_feature(Feature::Personality)
.enable_feature(Feature::FastMode)
.disable_feature(Feature::ShellSnapshot)
.with_root_config("chatgpt_base_url = \"http://127.0.0.1:1234\"")
.with_provider_config("requires_openai_auth = true")
@@ -28,7 +28,7 @@ fn mock_responses_config_composes_model_provider_features_and_extra_tables() {
"chatgpt_base_url = \"http://127.0.0.1:1234\"",
"model_provider = \"openai-custom\"",
"shell_snapshot = false",
"personality = true",
"fast_mode = true",
"[model_providers.openai-custom]\nname = \"OpenAI\"",
"base_url = \"http://127.0.0.1:1234/api/codex\"",
"requires_openai_auth = true",
@@ -44,10 +44,7 @@ fn legacy_mock_responses_writer_preserves_provider_auth_and_feature_overrides()
write_mock_responses_config_toml(
home.path(),
"http://127.0.0.1:1234",
&BTreeMap::from([
(Feature::Personality, true),
(Feature::ShellSnapshot, false),
]),
&BTreeMap::from([(Feature::FastMode, true), (Feature::ShellSnapshot, false)]),
/*auto_compact_limit*/ 321,
Some(true),
"openai",
@@ -62,7 +59,7 @@ fn legacy_mock_responses_writer_preserves_provider_auth_and_feature_overrides()
"model_auto_compact_token_limit = 321",
"openai_base_url = \"http://127.0.0.1:1234/v1\"",
"shell_snapshot = false",
"personality = true",
"fast_mode = true",
"[model_providers.openai]\nname = \"OpenAI\"",
"supports_websockets = false",
"requires_openai_auth = true",