live: heartbeat sync 2026-10-18T16:47:59-07:00 (Refs #11)

agent/memories/sharpshooter/frank-model-bench-1f78lrpnd0uoh/architecture.md

@@ -0,0 +1,13 @@
+## Runtime Stability
+- System runs must sustain operation beyond the first second; early termination within ~1s is a failure condition.
+
+## Performance Analysis
+- Performance gap analysis includes archon's KV quantization as a potential factor.
+- Evaluation uses index-lite metric for Qwen3.8-Flash-Next.
+- Composite index scores should be calculated and compared alongside individual model scores.
+
+## Reference Baselines
+- Agentic fine-tune evaluation: Qwen3-6-27B serves as the primary comparison baseline.
+- General capability evaluation: Qwen3.6 27B is to be evaluated against Claude Haiku 5-5.
+- Benchmark reference comparison target: qwen-flash-next.
+- In fair fight comparison, archon underperforms relative to reference.

agent/memories/sharpshooter/frank-model-bench-1f78lrpnd0uoh/product.md

@@ -0,0 +1,20 @@
+## Design and Testing Phases
+- Testing workflow supports pausing and resuming.
+- Test frank-archon after design implementation.
+- Adversarial reviews are conducted via the frank-archon subagent with extreme thoroughness.
+
+## Dashboard Access
+- Dashboard results accessible at http://127.0.0.1:8787
+
+## Benchmarking Workflow
+- When harness indicates failure or configuration changes (e.g., KV quantization), discard current benchmarking results and restart the entire benchmarking run.
+
+## Judge Selection
+- Use frank-zealot as the default judge for SciCode benchmark evaluation.
+- Judge must be performant enough for AA-LCR and HLE tasks.
+
+## Reference Comparison
+- Evaluate model performance against alternatives as part of the evaluation scope.
+- Reference comparison target is qwen-flash-next.
+- In fair fight comparison, archon underperforms relative to reference.
+- Score comparison to other models is included in the report.

agent/memories/sharpshooter/frank-model-bench-1f78lrpnd0uoh/state.json

@@ -0,0 +1,10 @@
+{
+	"v": 1,
+	"lastConsolidatedAt": 1791969871087,
+	"lastResult": {
+		"at": 1791852389828,
+		"sessions": 1,
+		"deltas": 1,
+		"model": "frank-zealot"
+	}
+}