werkrevisor reviews werkkompass #110
No Assignees
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: mi/werkrevisor#110
Reference in New Issue
Block a user
Host config for mi/werkkompass.
See docs/prs/.
🤖 Generated with Claude Code
🟢 werkrevisor GREEN for
df54aa6, superseded by the result for4aadc72🟢 werkrevisor: GREEN / PASS
Gate: approved by werkrevisor; merge waits for 2 open hint(s): reply to each line comment and resolve it, then restart the review or push a commit.
Change kind:
mixed.Reasons
Findings
Missing Test Coverage at
tests/test_config_budget_redact.py:35-46, severity unknown, counted as medium, hint, as a line commentMissing Test Coverage at
tests/test_config_budget_redact.py:124-130, severity unknown, counted as medium, hint, as a line commentClassifier
5 of 5 units passed all questions (1 PR document(s)).
Usage: model `openrouter/minimax/minimax-m3`, LLM done, tokens counted before the call 20776, calls 1, tokens in 20099 / out 4897 (reasoning included in out), duration 68.2 s, cost 0.0119 USD (reported), classifier `jev-1.13.0` calls 5, tokens in 3539, cost 0.0001 USD, finding judge `jev-1.13.0` calls 2, tokens in 3855, cost 0.0002 USD, change summary `openrouter/minimax/minimax-m3` via Minimax calls 1, tokens in 4898 / out 1028, cost 0.0027 USD (reported), PR document check `jev-1.13.0` calls 5, tokens in 6374, cost 0.0003 USD, build file judge `jev-1.13.0` calls 4, tokens in 5525, cost 0.0002 USD, phases: path_judge 0.4 s, classifier 0.5 s, change_summary 10.7 s, doc_check 0.5 s, context 0.6 s, model_review 54.9 s, finding_judge 0.3 s, config: defaults + host + repository (.werkrevisor.toml at main85f8b95).Reviewed head `df54aa6a4f7ea47ec40a604f19f7c73cc31c27d3` against base `85f8b95b546244edbf147e371022b7b6a6e9d96c`, werkrevisor 0.42.2.
werkrevisor: GREEN / PASS for
df54aa6a4f7ea47ec40a604f19f7c73cc31c27d3.The result comment of werkrevisor holds the details.
@@ -31,5 +31,5 @@def test_example_full_files_leave_room_for_the_diff_of_a_pr_at_the_size_limit(self):for name in ("werkprobe.toml", "werkrevisor.toml", "hsadmin-ng.toml"):with self.subTest(name=name):cfg = load_config(os.path.join(ROOT, "config", name))diff_chars = cfg.size.max_changed_lines * 80Missing Test Coverage, severity unknown, counted as medium (werkrevisor)
hint; Jev: real 0.75, severity 2.0 of 3
The new
werkkompass.tomlis added to only two of the five per-example test methods. The other three methods (test_example_full_files_leave_room_for_the_diff_of_a_pr_at_the_size_limit,test_examples_leave_the_countable_writing_rules_to_werkrevisor,test_examples_ask_for_a_severity_per_finding) check invariants the new config also satisfies (full-files sizing, the countable-writing-rules phrasing inextra_instructions, and the severity header /orange_min_medium = 2), so a future change to those invariants would slip past CI for werkkompass until production use reveals the regression. Adding the config to all three loops would keep the test suite uniformly strict across example configs.@@ -120,5 +120,5 @@self.assertEqual(example_config(name).review.orange_min_medium, 2)def test_hsadmin_ng_budget(self):self.assertEqual(load_config(os.path.join(ROOT, "config", "hsadmin-ng.toml")).budget.max_usd_per_review, 1.0)Missing Test Coverage, severity unknown, counted as medium (werkrevisor)
hint; Jev: real 0.71, severity 2.0 of 3
Same pattern as above: the loop in
test_examples_ask_for_a_severity_per_findingenumerates the seven existing configs but skipswerkkompass.toml. The config setsorange_min_medium = 2and includes the required severity phrasing inextra_instructions, so the test would pass if the config were added. Leaving it out weakens uniform enforcement of the severity-per-finding contract for the new repository.New commits pushed, approval review dismissed automatically according to repository settings
@@ -34,7 +34,7 @@ class ConfigTest(unittest.TestCase):def test_example_full_files_leave_room_for_the_diff_of_a_pr_at_the_size_limit(self):Fixed in
4aadc72: werkkompass.toml is in all five example-config loops now.@@ -122,7 +122,7 @@ class ConfigTest(unittest.TestCase):def test_examples_ask_for_a_severity_per_finding(self):for name in ("werkprobe.toml", "werkrevisor.toml", "werkator.toml", "werkcrappit.toml", "werkregie.toml",Fixed in
4aadc72: werkkompass.toml is in all five example-config loops now.🟢 werkrevisor: GREEN / PASS
Change kind:
mixed.Reasons
Classifier
5 of 5 units passed all questions (1 PR document(s)).
Usage: model `openrouter/minimax/minimax-m3`, LLM done, tokens counted before the call 22206, calls 1, tokens in 21436 / out 6214 (reasoning included in out), duration 67.3 s, cost 0.0139 USD (reported), classifier `jev-1.13.0` calls 5, tokens in 3539, cost 0.0001 USD, change summary `openrouter/minimax/minimax-m3` via Minimax calls 1, tokens in 5274 / out 911, cost 0.0026 USD (reported), PR document check `jev-1.13.0` calls 5, tokens in 6269, cost 0.0003 USD, build file judge `jev-1.13.0` calls 1, tokens in 1103, cost 0.0000 USD, phases: path_judge 0.3 s, classifier 0.5 s, change_summary 8.8 s, doc_check 0.5 s, context 0.3 s, model_review 56.5 s, config: defaults + host + repository (.werkrevisor.toml at main85f8b95).Reviewed head `4aadc72d1a100f710c8e2a9d266d253676d587e1` against base `85f8b95b546244edbf147e371022b7b6a6e9d96c`, werkrevisor 0.42.2.
werkrevisor: GREEN / PASS for
4aadc72d1a100f710c8e2a9d266d253676d587e1.The result comment of werkrevisor holds the details.
Codex review: critical
The new host configuration controls human-approval requirements and enables automatic merging for werkkompass. A mistake could allow unsafe changes to merge without human review, which the configuration tests would not establish.
No findings.
The merge waits for the status
Model `gpt-6-astra`, effort high, tokens in 203310 (cached 153088) / out 1028, 55.4 s, 6 commands; reviewed head `4aadc72d1a100f710c8e2a9d266d253676d587e1`.ai-review/codex: every line comment resolved, and the approval of a human for a critical PR or for a finding that is only here.