Skip to content

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

fix(proxy): preserve Responses memory continuations with store=false (#1103)

Ian Ker-Seymercdfeeacc63e6
diff --git a/CHANGELOG.md b/CHANGELOG.md
index bac7870d8..08a561473 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -29,6 +29,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
### Bug Fixes
+* **proxy:** force Responses API `store=true` when Headroom injects memory tools so `previous_response_id` continuations work after memory tool calls from clients that requested `store=false` ([#1103](https://github.com/chopratejas/headroom/pull/1103)).
* **proxy:** build SSL contexts for custom CA bundles so enterprise/private PKI roots work with Python/OpenSSL strict verification.
* **proxy:** route Codex OAuth image generation and edit requests through the ChatGPT Codex image backend, while preserving OpenAI API-key image passthrough ([#1215](https://github.com/chopratejas/headroom/pull/1215)).
* **wrap (codex):** keep RTK guidance in the global Codex `AGENTS.md` instead of modifying the shared project `AGENTS.md` ([#1235](https://github.com/chopratejas/headroom/issues/1235)).
diff --git a/headroom/proxy/handlers/openai.py b/headroom/proxy/handlers/openai.py
index 96f0ca0f1..d61ced6ee 100644
--- a/headroom/proxy/handlers/openai.py
+++ b/headroom/proxy/handlers/openai.py
@@ -321,6 +321,29 @@ def _compact_openai_responses_tools(
return updated, True, before, after
+def _ensure_responses_store_for_memory_tools(
+ payload: dict[str, Any],
+ *,
+ memory_tools_injected: bool,
+) -> bool:
+ """Keep Responses API memory-tool continuations addressable.
+
+ Memory tools are transparent to clients: Headroom executes the emitted
+ function_call, then sends function_call_output in a continuation request
+ using previous_response_id. OpenAI only allows that continuation when the
+ previous response was stored. Clients such as pi/Codex can set store=false
+ to avoid retaining ordinary responses, but that makes memory-tool
+ continuations fail with previous_response_not_found.
+
+ Return True when this function changes the payload.
+ """
+
+ if memory_tools_injected and payload.get("store") is False:
+ payload["store"] = True
+ return True
+ return False
+
+
def _responses_input_item_text_bytes(item: Any) -> int:
if not isinstance(item, dict):
return _json_byte_len(item)
@@ -3093,6 +3116,14 @@ class OpenAIHandlerMixin:
if mem_tools_injected:
body["tools"] = resp_tools
logger.info(f"[{request_id}] Memory: Injected memory tools (openai/responses)")
+
+ if _ensure_responses_store_for_memory_tools(
+ body,
+ memory_tools_injected=True,
+ ):
+ logger.info(
+ f"[{request_id}] Memory: forced store=true for Responses memory tool continuation"
+ )
except Exception as e:
logger.warning(f"[{request_id}] Memory injection failed (responses): {e}")
elif self.memory_handler and memory_user_id and _bypass:
diff --git a/tests/test_openai_responses_context_compaction.py b/tests/test_openai_responses_context_compaction.py
index 264d693a3..6f5eff3ec 100644
--- a/tests/test_openai_responses_context_compaction.py
+++ b/tests/test_openai_responses_context_compaction.py
@@ -6,6 +6,7 @@ from typing import Any
from headroom.proxy.handlers.openai import (
OpenAIHandlerMixin,
_compact_openai_responses_tools,
+ _ensure_responses_store_for_memory_tools,
_openai_responses_context_budget,
)
from headroom.transforms.content_router import (
@@ -435,3 +436,50 @@ def test_content_router_retries_kompress_when_structured_strategy_noops(monkeypa
assert compressed_tokens == 2
# The fallback chain must record both strategies it tried.
assert strategy_chain == ["smart_crusher", "kompress"]
+
+
+def test_responses_memory_tools_store_false_regression() -> None:
+ """Regression: store=false makes previous_response_id continuations fail."""
+
+ payload = {"model": "gpt-5.5", "input": "remember this", "store": False}
+
+ changed = _ensure_responses_store_for_memory_tools(
+ payload,
+ memory_tools_injected=True,
+ )
+
+ assert changed is True
+ assert payload["store"] is True
+
+
+def test_responses_memory_tools_do_not_change_unrelated_requests() -> None:
+ no_memory_payload = {"model": "gpt-5.5", "input": "plain", "store": False}
+ already_stored_payload = {"model": "gpt-5.5", "input": "plain", "store": True}
+ default_store_payload = {"model": "gpt-5.5", "input": "plain"}
+
+ assert (
+ _ensure_responses_store_for_memory_tools(
+ no_memory_payload,
+ memory_tools_injected=False,
+ )
+ is False
+ )
+ assert no_memory_payload["store"] is False
+
+ assert (
+ _ensure_responses_store_for_memory_tools(
+ already_stored_payload,
+ memory_tools_injected=True,
+ )
+ is False
+ )
+ assert already_stored_payload["store"] is True
+
+ assert (
+ _ensure_responses_store_for_memory_tools(
+ default_store_payload,
+ memory_tools_injected=True,
+ )
+ is False
+ )
+ assert "store" not in default_store_payload