applyPromptTypeParams(request, "docnotfound");
        request.setStream(true);

        streamChatWithConcurrencyControl(request, callback);
    }

    @Override
    public void generateSummaryResponse(final String userMessage, final List<Map<String, Object>> documents, final List<LlmMessage> history,
            final LlmStreamCallback callback) {
        final LlmChatRequest request = new LlmChatRequest();

        }

        final String result = client.testBuildContext(documents);
        assertTrue(result.length() <= 300); // some overhead from header text
    }

    // ========== P6: generateSummaryResponse size limit and HTML stripping tests ==========

    @Test
    public void test_generateSummaryResponse_stripsHtmlFromContent() {
        client.setTestContextMaxChars(10000);

                    phaseStartTime = System.currentTimeMillis();
                    callback.onPhaseStart(ChatPhaseCallback.PHASE_ANSWER, "Generating summary...");
                    llmClientManager.generateSummaryResponse(userMessage, fullDocs, history, (chunk, done) -> {
                        fullResponse.append(chunk);
                        callback.onChunk(chunk, done);
                    });