sususu f0034ca663 fix(claude): restore cloaked prompt-cache ownership and native shape ...

Cloaked Responses/Chat/Gemini→Claude traffic was skipping ensureCacheControl
once the first-user cloak marker existed, freezing multi-turn cache hits.
Always ensure section-independent system and rolling-message breakpoints for
non-native callers, keep confirmed Claude Code placement intact, and leave
tools unstamped when a usable system prompt already covers the prefix.

Align the wire shape with the Claude Code 2.1.220/2.1.221/2.1.227 cache-control
constructor: the default is {"type":"ephemeral"} with no ttl. A 1h body ttl is
applied only after placement, only for OAuth credentials, and only onto blocks
that already carry cache_control, so it stays strictly paired with
extended-cache-ttl-2025-04-11. Explicit caller-owned 5m choices survive.

Match the native rolling selector (skip thinking-like assistant tails; final
system string special case), fall back to a tools breakpoint when system is
absent or blank, and leave empty/whitespace string systems unconverted so they
cannot double-stamp tools plus a useless system host.

2026-08-11 19:33:40 +08:00

..

helps

fix(antigravity): harden per-credential transport pooling

2026-08-11 18:33:50 +08:00

aistudio_executor_test.go

fix(thinking): close summary translation gaps

2026-07-30 23:26:20 +08:00

aistudio_executor.go

fix(thinking): honor normalized summary payloads

2026-07-31 12:45:13 +08:00

antigravity_executor_auth.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

antigravity_executor_buildrequest_test.go

feat(runtime): implement derived session identity features and tests

2026-07-26 03:07:21 +08:00

antigravity_executor_credits_test.go

test(antigravity): clear credits state in place to fix a cleanup data race

2026-08-11 09:57:15 +08:00

antigravity_executor_credits.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

antigravity_executor_execute.go

perf(antigravity): translate each upstream request once

2026-08-11 16:07:42 +08:00

antigravity_executor_interactions_test.go

feat(interactions): add support for Google Interactions

2026-07-07 02:39:13 +08:00

antigravity_executor_keepalive_test.go

perf(antigravity): reuse native upstream connections

2026-08-11 18:33:50 +08:00

antigravity_executor_request.go

perf(antigravity): reuse native upstream connections

2026-08-11 18:33:50 +08:00

antigravity_executor_signature_test.go

feat(antigravity): obfuscate sensitive words in system instructions

2026-08-05 00:27:32 +08:00

antigravity_executor_stream.go

perf(antigravity): translate each upstream request once

2026-08-11 16:07:42 +08:00

antigravity_executor_tokens.go

perf(antigravity): reuse native upstream connections

2026-08-11 18:33:50 +08:00

antigravity_executor_transport_test.go

fix(antigravity): harden per-credential transport pooling

2026-08-11 18:33:50 +08:00

antigravity_executor.go

fix(antigravity): harden per-credential transport pooling

2026-08-11 18:33:50 +08:00

antigravity_preupstream_rewrite_differential_test.go

perf(antigravity): batch pre-upstream JSON rewrites

2026-08-11 16:07:42 +08:00

antigravity_preupstream_rewrite_legacy_oracle_test.go

perf(antigravity): batch pre-upstream JSON rewrites

2026-08-11 16:07:42 +08:00

antigravity_reasoning_replay_clear_test.go

fix(antigravity): preserve complete reasoning replay (#4525)

2026-07-25 14:07:31 +08:00

antigravity_reasoning_replay_index_test.go

refactor(antigravity): route replay merge through the request index

2026-08-11 09:57:15 +08:00

antigravity_reasoning_replay_legacy_oracle_test.go

refactor(antigravity): route replay merge through the request index

2026-08-11 09:57:15 +08:00

antigravity_reasoning_replay_test.go

refactor(antigravity): route replay merge through the request index

2026-08-11 09:57:15 +08:00

antigravity_reasoning_replay.go

perf(antigravity): batch pre-upstream JSON rewrites

2026-08-11 16:07:42 +08:00

antigravity_refresh_test.go

fix(antigravity): harden per-credential transport pooling

2026-08-11 18:33:50 +08:00

antigravity_schema_sanitize_test.go

fix(antigravity): strip propertyNames inside a property named "properties"

2026-08-03 16:09:22 +08:00

caching_verify_test.go

fix(claude): restore cloaked prompt-cache ownership and native shape

2026-08-11 19:33:40 +08:00

claude_executor_auth_race_test.go

fix(claude): replay measured OAuth wire, Fast and diagnostic profiles

2026-08-03 14:47:26 +08:00

claude_executor_auth_test.go

fix(claude): harden request lifecycle

2026-08-03 14:47:26 +08:00

claude_executor_auth.go

fix(claude): harden OAuth identity and native routing

2026-08-03 14:47:26 +08:00

claude_executor_beta_policy_test.go

fix(claude): pair diagnostics with cache beta

2026-08-03 15:44:44 +08:00

claude_executor_cloaking.go

fix(claude): restore cloaked prompt-cache ownership and native shape

2026-08-11 19:33:40 +08:00

claude_executor_diagnostics_test.go

fix(claude): pair diagnostics with cache beta

2026-08-03 15:44:44 +08:00

claude_executor_diagnostics.go

fix(claude): harden OAuth identity and native routing

2026-08-03 14:47:26 +08:00

claude_executor_execute.go

fix(claude): restore cloaked prompt-cache ownership and native shape

2026-08-11 19:33:40 +08:00

claude_executor_fast_error_test.go

fix(claude): harden request lifecycle

2026-08-03 14:47:26 +08:00

claude_executor_fast_error.go

fix(claude): harden request lifecycle

2026-08-03 14:47:26 +08:00

claude_executor_request_bench_test.go

perf(claude): batch OAuth tool name rewrites

2026-08-03 14:47:26 +08:00

claude_executor_request_remap_test.go

fix(claude): recover malformed OAuth MCP aliases when reverse-remapping Claude tool names

2026-08-10 04:32:28 +08:00

claude_executor_request.go

fix(claude): restore cloaked prompt-cache ownership and native shape

2026-08-11 19:33:40 +08:00

claude_executor_stream.go

fix(claude): restore cloaked prompt-cache ownership and native shape

2026-08-11 19:33:40 +08:00

claude_executor_test.go

fix(claude): restore cloaked prompt-cache ownership and native shape

2026-08-11 19:33:40 +08:00

claude_executor_thinking_signature_test.go

test(claude): preserve adaptive thinking signatures

2026-08-03 14:47:26 +08:00

claude_executor_tokens.go

feat(compat): preserve compat-mode thinking/signature blocks for API-key models

2026-08-06 17:19:24 +08:00

claude_executor_wire_casing_test.go

fix(claude): send Anthropic header names with the real client's casing

2026-08-03 14:47:26 +08:00

claude_executor.go

feat(compat): preserve compat-mode thinking/signature blocks for API-key models

2026-08-06 17:19:24 +08:00

claude_signing_test.go

feat(claude): align OAuth wire identity and TLS with Claude Code 2.1.220

2026-08-03 14:47:26 +08:00

claude_signing.go

feat(claude): align OAuth wire identity and TLS with Claude Code 2.1.220

2026-08-03 14:47:26 +08:00

claude_thinking_replay_test.go

fix(claude): add Claude-compatible thinking replay persistence for multi-turn sessions

2026-08-09 04:36:39 +08:00

claude_thinking_replay.go

fix(claude): add Claude-compatible thinking replay persistence for multi-turn sessions

2026-08-09 04:36:39 +08:00

codex_executor_auth.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

codex_executor_cache_test.go

feat(runtime): implement derived session identity features and tests

2026-07-26 03:07:21 +08:00

codex_executor_compact_test.go

feat(executor, handlers): enhance websocket output restoration and compact compaction logic

2026-07-11 11:42:51 +08:00

codex_executor_execute.go

feat(compat): preserve compat-mode thinking/signature blocks for API-key models

2026-08-06 17:19:24 +08:00

codex_executor_imagegen_test.go

feat(executor): add normalization for parallel tool calls in Codex executors

2026-07-18 03:58:03 +08:00

codex_executor_input_ids_test.go

feature(codex): Normalize invalid sub2api message item IDs during Codex input sanitization

2026-08-03 02:45:49 +08:00

codex_executor_instructions_test.go

chore: upgrade CLIProxyAPI dependency to v7 across the project

2026-05-08 11:46:46 +08:00

codex_executor_parallel_tool_calls_test.go

feat(executor): add normalization for parallel tool calls in Codex executors

2026-07-18 03:58:03 +08:00

codex_executor_reasoning_replay_cache_test.go

perf(codex): make replay prefix fingerprints incremental

2026-07-24 11:10:58 -07:00

codex_executor_reasoning.go

Merge PR #4546 into dev

2026-07-26 19:11:32 +08:00

codex_executor_request.go

feat(compat): preserve compat-mode thinking/signature blocks for API-key models

2026-08-06 17:19:24 +08:00

codex_executor_retry_test.go

feat(executor): handle usage limit errors and enhance retry logic

2026-06-17 07:19:02 +08:00

codex_executor_signature_test.go

fix(executor): strip orphan reasoning ids when store is disabled (#4285)

2026-07-14 00:35:48 +08:00

codex_executor_spawn_agent_test.go

feat(codex): add model-level is-compat flag to rewrite MultiAgentV2 agent_message for Responses-compatible endpoints

2026-08-06 04:49:28 +08:00

codex_executor_stream_output_test.go

fix(codex): forward sequential cutoff reasoning summaries

2026-08-09 18:01:28 +08:00

codex_executor_stream.go

fix(codex): forward sequential cutoff reasoning summaries

2026-08-09 18:01:28 +08:00

codex_executor_terminal.go

fix(openai): expose only client-fault streaming errors

2026-08-06 20:49:23 +08:00

codex_executor_tokens.go

feat(compat): preserve compat-mode thinking/signature blocks for API-key models

2026-08-06 17:19:24 +08:00

codex_executor_translate_test.go

refactor(executor): consolidate Codex request translation logic

2026-05-31 14:38:54 +08:00

codex_executor.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

codex_openai_images_extract_test.go

perf(codex): avoid rebuilding completed JSON when extracting generated images

2026-06-08 15:47:14 +00:00

codex_openai_images_test.go

feat(codex): add disable-codex-cloaking config option and refine header management

2026-07-30 16:37:40 +08:00

codex_openai_images.go

fix(thinking): honor normalized summary payloads

2026-07-31 12:45:13 +08:00

codex_websockets_connection.go

Reduce Codex request amplification for large payloads

2026-08-07 19:16:06 +08:00

codex_websockets_errors.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

codex_websockets_execute.go

feat(codex): add model-level is-compat flag to rewrite MultiAgentV2 agent_message for Responses-compatible endpoints

2026-08-06 04:49:28 +08:00

codex_websockets_executor_store_test.go

feat: add Home credential concurrency support

2026-07-23 13:42:08 +08:00

codex_websockets_executor_test.go

Reduce Codex request amplification for large payloads

2026-08-07 19:16:06 +08:00

codex_websockets_executor.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

codex_websockets_request.go

fix(codex): align websocket cloaking headers

2026-08-01 13:01:33 +08:00

codex_websockets_session.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

codex_websockets_spawn_agent_test.go

feat(executor): replace sdktranslator.TranslateRequest with helps.TranslateRequestWithCodexMultiAgentV2

2026-07-25 00:32:37 +08:00

codex_websockets_stream.go

feat(codex): add model-level is-compat flag to rewrite MultiAgentV2 agent_message for Responses-compatible endpoints

2026-08-06 04:49:28 +08:00

executor_payload_optimization_test.go

perf(executor): replace sjson.SetBytes with optimized helpers for conditional payload updates

2026-07-20 13:57:14 +08:00

gemini_executor_test.go

feat(thinking): preserve cross-protocol summary visibility

2026-07-30 21:53:30 +08:00

gemini_executor.go

feat(compat): preserve compat-mode thinking/signature blocks for API-key models

2026-08-06 17:19:24 +08:00

gemini_vertex_executor.go

refactor(executor): consolidate thinking.ApplyThinking into helps.ApplyRequestThinking

2026-07-29 14:14:18 +08:00

home_codex_terminal_test.go

feat: add Home credential concurrency support

2026-07-23 13:42:08 +08:00

kimi_executor_test.go

fix(openai): preserve assistant content when converting Responses tool-call turns

2026-08-10 05:13:24 +08:00

kimi_executor.go

fix(kimi): canonicalize K2.7 Code model aliases to official Kimi-For-Coding IDs

2026-08-08 05:08:42 +08:00

kimi_thinking_replay_test.go

fix(kimi): canonicalize K2.7 Code model aliases to official Kimi-For-Coding IDs

2026-08-08 05:08:42 +08:00

kimi_thinking_replay.go

fix(claude): add Claude-compatible thinking replay persistence for multi-turn sessions

2026-08-09 04:36:39 +08:00

openai_compat_executor_compact_test.go

fix(openai): drop OpenAI stream chunks after [DONE]

2026-08-05 05:13:09 +08:00

openai_compat_executor_reasoning_test.go

feat(compat): preserve Claude thinking/tool-call content for is-compat OpenAI compatibility models

2026-08-07 06:26:33 +08:00

openai_compat_executor_tool_results_test.go

feat(executor): normalize OpenAI tool results for text-only compatibility models

2026-08-03 03:37:24 +08:00

openai_compat_executor.go

fix(cliproxy): delegate OpenAI-compatible OAuth refresh to plugin auth providers

2026-08-08 06:08:23 +08:00

openai_responses_signature_test.go

Reduce Codex request amplification for large payloads

2026-08-07 19:16:06 +08:00

openai_responses_signature.go

Reduce Codex request amplification for large payloads

2026-08-07 19:16:06 +08:00

websocket_lifecycle_bind_test.go

feat: add Home credential concurrency support

2026-07-23 13:42:08 +08:00

websocket_session_target_test.go

feat: add Home credential concurrency support

2026-07-23 13:42:08 +08:00

xai_executor_auth.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

xai_executor_execute.go

fix(responses): preserve original request model on response.created/response.in_progress payloads

2026-08-04 18:37:01 +08:00

xai_executor_media.go

fix(executor): add failure tracking and reporting for video execution

2026-07-27 00:01:49 +08:00

xai_executor_request.go

fix(xai): normalize forced web_search tool_choice during Responses request prep

2026-08-09 23:15:46 +08:00

xai_executor_response.go

fix(xai): remap bad-credentials 403s to unauthorized

2026-08-07 06:48:09 +08:00

xai_executor_stream.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

xai_executor_test.go

fix(xai): normalize forced web_search tool_choice during Responses request prep

2026-08-09 23:15:46 +08:00

xai_executor_tokens.go

chore(pluginhost): refactor and remove unused interceptors and executor methods

2026-07-26 14:31:45 +08:00

xai_executor.go

feat(xai): bump client version to 0.2.120 and include Grok Shell auth headers on XAI requests

2026-08-05 07:28:55 +08:00

xai_reasoning_replay.go

fix(xai): clear replay after compaction

2026-07-12 23:59:39 +08:00

xai_status_err_test.go

fix(xai): remap bad-credentials 403s to unauthorized

2026-08-07 06:48:09 +08:00

xai_websockets_executor_test.go

fix(xai): remap bad-credentials 403s to unauthorized

2026-08-07 06:48:09 +08:00

xai_websockets_executor.go

fix(xai): remap bad-credentials 403s to unauthorized

2026-08-07 06:48:09 +08:00