diff --git a/evals/reports/2026-05-25-live-drift.yaml b/evals/reports/2026-05-25-live-drift.yaml index 0f515b0..2b0220a 100644 --- a/evals/reports/2026-05-25-live-drift.yaml +++ b/evals/reports/2026-05-25-live-drift.yaml @@ -6,7 +6,7 @@ eval_id: live-profile-drift profile: cto-planb status: pass score: 100 -checked_at: '2026-05-25T17:14:09Z' +checked_at: '2026-05-25T17:20:57Z' checks: correctness: pass verification: pass @@ -76,7 +76,7 @@ commands: - command: hermes -p cto-planb skills list cwd: /home/svrnty/workspaces/hermes returncode: 0 - duration_ms: 211 + duration_ms: 235 stdout: " Installed Skills \n\u250F\ \u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\ \u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2533\u2501\ @@ -113,7 +113,7 @@ commands: - command: hermes -p cto-planb mcp list cwd: /home/svrnty/workspaces/hermes returncode: 0 - duration_ms: 401 + duration_ms: 470 stdout: "\n MCP Servers:\n\n Name Transport Tools\ \ Status \n \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\ \u2500\u2500\u2500\u2500\u2500\u2500 \u2500\u2500\u2500\u2500\u2500\u2500\u2500\ diff --git a/evals/reports/2026-05-25-live-promotion-readiness.yaml b/evals/reports/2026-05-25-live-promotion-readiness.yaml index bf8ed3f..200a645 100644 --- a/evals/reports/2026-05-25-live-promotion-readiness.yaml +++ b/evals/reports/2026-05-25-live-promotion-readiness.yaml @@ -59,7 +59,7 @@ eval_results: command: command: hermes -p cto-planb skills list returncode: 0 - duration_ms: 232 + duration_ms: 210 stdout: " Installed Skills \n\u250F\ \u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\ \u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2501\u2533\u2501\ @@ -100,7 +100,7 @@ eval_results: command: command: hermes -p cto-planb mcp list returncode: 0 - duration_ms: 397 + duration_ms: 443 stdout: "\n MCP Servers:\n\n Name Transport \ \ Tools Status \n \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\ \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500 \u2500\u2500\u2500\u2500\u2500\ diff --git a/evals/reports/2026-05-25-local-regression-execution-slice.yaml b/evals/reports/2026-05-25-local-regression-execution-slice.yaml index c81c30f..f662e24 100644 --- a/evals/reports/2026-05-25-local-regression-execution-slice.yaml +++ b/evals/reports/2026-05-25-local-regression-execution-slice.yaml @@ -38,19 +38,19 @@ eval_results: - cto/evals/reports/2026-05-25-promotion-fixture-execution.yaml command: python3 evals/runners/run-promotion-fixtures.py --output evals/reports/2026-05-25-promotion-fixture-execution.yaml --artifact-output evals/artifacts/2026-05-25-promotion-fixture-execution.json - duration_ms: 739 + duration_ms: 744 - eval_id: live-promotion-readiness status: pass evidence: - cto/evals/reports/2026-05-25-live-promotion-readiness.yaml command: python3 evals/runners/run-live-promotion-readiness.py --output evals/reports/2026-05-25-live-promotion-readiness.yaml - duration_ms: 668 + duration_ms: 693 - eval_id: static-prd-contract status: pass evidence: - tests/e2e/test_j_cto_webui_prd.py command: pytest -q tests/e2e/test_j_cto_webui_prd.py - duration_ms: 1198 + duration_ms: 1216 - eval_id: webui-cto-event-browser status: pass evidence: @@ -58,32 +58,33 @@ eval_results: - hermes-webui/tests/test_cancel_interrupt.py command: pytest -q tests/test_cto_events.py tests/test_live_tool_callback_events.py tests/test_cto_webui_journal_e2e.py tests/test_cto_browser_e2e.py tests/test_cancel_interrupt.py - duration_ms: 3090 + tests/test_approval_queue.py + duration_ms: 2364 - eval_id: webui-cto-live-streaming status: pass evidence: - hermes-webui/tests/test_cto_live_streaming_e2e.py command: pytest -q tests/test_cto_live_streaming_e2e.py - duration_ms: 1906 + duration_ms: 1220 - eval_id: live-profile-drift status: pass evidence: - cto/evals/reports/2026-05-25-live-drift.yaml command: python3 evals/runners/drift.py --output evals/reports/2026-05-25-live-drift.yaml - duration_ms: 661 + duration_ms: 752 - eval_id: eval-report-scoring status: pass evidence: - cto/evals/reports/*.yaml command: bash -lc for r in evals/reports/*.yaml; do python3 evals/runners/score.py "$r"; done - duration_ms: 289 + duration_ms: 319 - eval_id: diff-whitespace-check status: pass evidence: - git diff --check command: git diff --check - duration_ms: 6 + duration_ms: 4 commands: - command: python3 evals/runners/run-promotion-suite.py --output evals/reports/2026-05-25-promotion-suite-readiness.yaml cwd: /home/svrnty/workspaces/hermes/cto @@ -97,7 +98,7 @@ commands: --artifact-output evals/artifacts/2026-05-25-promotion-fixture-execution.json cwd: /home/svrnty/workspaces/hermes/cto returncode: 0 - duration_ms: 739 + duration_ms: 744 stdout: 'wrote /home/svrnty/workspaces/hermes/cto/evals/reports/2026-05-25-promotion-fixture-execution.yaml wrote /home/svrnty/workspaces/hermes/cto/evals/artifacts/2026-05-25-promotion-fixture-execution.json @@ -107,7 +108,7 @@ commands: - command: python3 evals/runners/run-live-promotion-readiness.py --output evals/reports/2026-05-25-live-promotion-readiness.yaml cwd: /home/svrnty/workspaces/hermes/cto returncode: 0 - duration_ms: 668 + duration_ms: 693 stdout: 'wrote evals/reports/2026-05-25-live-promotion-readiness.yaml ' @@ -115,38 +116,39 @@ commands: - command: pytest -q tests/e2e/test_j_cto_webui_prd.py cwd: /home/svrnty/workspaces/hermes returncode: 0 - duration_ms: 1198 + duration_ms: 1216 stdout: '.......... [100%] - 10 passed in 1.02s + 10 passed in 1.04s ' stderr: '' - command: pytest -q tests/test_cto_events.py tests/test_live_tool_callback_events.py tests/test_cto_webui_journal_e2e.py tests/test_cto_browser_e2e.py tests/test_cancel_interrupt.py + tests/test_approval_queue.py cwd: /home/svrnty/workspaces/hermes/hermes-webui returncode: 0 - duration_ms: 3090 - stdout: '...................... [100%] + duration_ms: 2364 + stdout: '...................................... [100%] - 22 passed in 2.63s + 38 passed in 1.89s ' stderr: '' - command: pytest -q tests/test_cto_live_streaming_e2e.py cwd: /home/svrnty/workspaces/hermes/hermes-webui returncode: 0 - duration_ms: 1906 + duration_ms: 1220 stdout: '. [100%] - 1 passed in 1.49s + 1 passed in 0.88s ' stderr: '' - command: python3 evals/runners/drift.py --output evals/reports/2026-05-25-live-drift.yaml cwd: /home/svrnty/workspaces/hermes/cto returncode: 0 - duration_ms: 661 + duration_ms: 752 stdout: 'wrote evals/reports/2026-05-25-live-drift.yaml ' @@ -155,7 +157,7 @@ commands: "$r"; done cwd: /home/svrnty/workspaces/hermes/cto returncode: 0 - duration_ms: 289 + duration_ms: 319 stdout: 'ok ok @@ -181,7 +183,7 @@ commands: - command: git diff --check cwd: /home/svrnty/workspaces/hermes returncode: 0 - duration_ms: 6 + duration_ms: 4 stdout: '' stderr: '' notes: diff --git a/evals/runners/run-local-regression.py b/evals/runners/run-local-regression.py index f3e8efe..4210dbc 100755 --- a/evals/runners/run-local-regression.py +++ b/evals/runners/run-local-regression.py @@ -163,6 +163,7 @@ def build_report(output: Path) -> dict[str, Any]: "tests/test_cto_webui_journal_e2e.py", "tests/test_cto_browser_e2e.py", "tests/test_cancel_interrupt.py", + "tests/test_approval_queue.py", ], cwd=WEBUI_ROOT, timeout=180,