werkrevisor 0.31.6: hints on PR documents are line comments again #72

2026-10-10T12:25:42+02:00

claude-code commented

2026-10-10 12:25:42 +02:00

🤖 Generated with Claude Code

See [the PR document](docs/prs/2026-10-10-PR%2372-pr-doc-hints-as-line-comments.md). 🤖 Generated with [Claude Code](https://claude.com/claude-code)

claude-code added 1 commit 2026-10-10 12:25:43 +02:00

PR document: hints on PR documents are line comments again

werkator build succeeded after 00:31

Details

werkator/werkrevisor-statistics build succeeded after 00:55

Details

184fe786f9

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

mi added 1 commit 2026-10-10 12:27:27 +02:00

Hints on PR documents are line comments again; tools/review-rounds counts the rounds (0.31.6)

werkator build succeeded after 00:27

Details

werkator/werkrevisor-statistics build succeeded after 01:17

Details

ai-review/assessment AI findings; a developer approval after checking them unlocks the merge; review requested from mi; merge scheduled for when the checks succeed

Details

ai-review/gate build succeeded after 01:26

Details

715f1bab61

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

mi force-pushed claude-code/pr-doc-hints-as-line-comments from 753b032c3a to 715f1bab61

2026-10-10 12:27:27 +02:00

Compare

claude-code marked the pull request as ready for review 2026-10-10 12:30:05 +02:00

ai-reviewer commented

2026-10-10 12:31:33 +02:00

🟠 werkrevisor ORANGE for 715f1ba, superseded by the result for fb3f5b3

🟠 werkrevisor: ORANGE / FINDINGS

Change kind: mixed.

Reasons

ORANGE 1 finding(s) that Jev rates as probably real and severe: tools/review-rounds:32
YELLOW spec: docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md Scenario#72.02 hard_to_read (0.51 > yellow_above = 0.5)
hint 5 changed file(s) that neither weaken the checks nor touch security according to Jev
hint 1 build change(s) of low risk according to Jev: statistics/build.gradle.kts (weakens checks 0.02, risk 0.00, security 0.01)
hint 7 section(s) of the PR document agree with the change summary of openrouter/minimax/minimax-m3 according to Jev
hint 1 finding(s) of low severity or rated minor by Jev: docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md:21

Findings

API Pagination Limit at tools/review-rounds:32, severity unknown, counted as medium, ORANGE, as a line comment
Ambiguous Table Row at docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md:21, severity unknown, counted as medium

hint; Jev: real 0.55, severity 1.1 of 3

The row | werkrevisor #70, two | ... uses a trailing "two" to indicate two findings on that PR. Readers can misread it as the PR number ("#70,2") or a duplicate entry. Spelling it as werkrevisor #70 (2) would carry the same information unambiguously.

Classifier

Unit	Question	Problem	Confidence	Verdict
docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md Scenario#72.02	hard_to_read	0.51	0.49	YELLOW

6 of 7 units passed all questions (1 PR document(s)).
1 answer ignored because of a confidence below [classifier] min_confidence.

_{Usage: model `openrouter/minimax/minimax-m3`, LLM done, tokens counted before the call 46581, calls 1, tokens in 45016 / out 7154 (reasoning included in out), duration 81.4 s, cost 0.0221 USD (reported), classifier `jev-1.13.0` calls 7, tokens in 5457, cost 0.0002 USD, finding judge `jev-1.13.0` calls 2, tokens in 2639, cost 0.0001 USD, change summary `openrouter/minimax/minimax-m3` via Minimax calls 1, tokens in 2348 / out 992, cost 0.0019 USD (reported), PR document check `jev-1.13.0` calls 7, tokens in 9548, cost 0.0004 USD, build file judge `jev-1.13.0` calls 6, tokens in 3969, cost 0.0002 USD, phases: path_judge 0.5 s, classifier 0.5 s, change_summary 9.4 s, doc_check 0.5 s, context 0.4 s, model_review 69.4 s, finding_judge 0.3 s, config: defaults + host + repository (.werkrevisor.toml at main 0323f38).

Reviewed head `715f1bab610dcd4d057139a12ec72fe9941c077f` against base `0323f38e4cf90ea3f21c341890cacbb40c4f19dd`, werkrevisor 0.31.5.}

<details><summary>🟠 werkrevisor ORANGE for <code>715f1ba</code>, superseded by <a href="http://git.javagil.de/mi/werkrevisor/pulls/72#issuecomment-3088">the result for <code>fb3f5b3</code></a></summary> ## 🟠 werkrevisor: ORANGE / FINDINGS Change kind: `mixed`. ### Reasons - **ORANGE** 1 finding(s) that Jev rates as probably real and severe: tools/review-rounds:32 - **YELLOW** spec: docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md Scenario#72.02 hard_to_read (0.51 > yellow_above = 0.5) - **hint** 5 changed file(s) that neither weaken the checks nor touch security according to Jev - **hint** 1 build change(s) of low risk according to Jev: statistics/build.gradle.kts (weakens checks 0.02, risk 0.00, security 0.01) - **hint** 7 section(s) of the PR document agree with the change summary of openrouter/minimax/minimax-m3 according to Jev - **hint** 1 finding(s) of low severity or rated minor by Jev: docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md:21 ### Findings 1. **API Pagination Limit** at `tools/review-rounds:32`, severity unknown, counted as medium, ORANGE, as a line comment 2. **Ambiguous Table Row** at `docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md:21`, severity unknown, counted as medium hint; Jev: real 0.55, severity 1.1 of 3 The row `| werkrevisor #70, two | ...` uses a trailing "two" to indicate two findings on that PR. Readers can misread it as the PR number ("#70,2") or a duplicate entry. Spelling it as `werkrevisor #70 (2)` would carry the same information unambiguously. ### Classifier | Unit | Question | Problem | Confidence | Verdict | |---|---|---|---|---| | docs/prs/2026-10-10-PR#72-pr-doc-hints-as-line-comments.md Scenario#72.02 | hard_to_read | 0.51 | 0.49 | YELLOW | 6 of 7 units passed all questions (1 PR document(s)). 1 answer ignored because of a confidence below [classifier] min_confidence. Usage: model `openrouter/minimax/minimax-m3`, LLM done, tokens counted before the call 46581, calls 1, tokens in 45016 / out 7154 (reasoning included in out), duration 81.4 s, cost 0.0221 USD (reported), classifier `jev-1.13.0` calls 7, tokens in 5457, cost 0.0002 USD, finding judge `jev-1.13.0` calls 2, tokens in 2639, cost 0.0001 USD, change summary `openrouter/minimax/minimax-m3` via Minimax calls 1, tokens in 2348 / out 992, cost 0.0019 USD (reported), PR document check `jev-1.13.0` calls 7, tokens in 9548, cost 0.0004 USD, build file judge `jev-1.13.0` calls 6, tokens in 3969, cost 0.0002 USD, phases: path_judge 0.5 s, classifier 0.5 s, change_summary 9.4 s, doc_check 0.5 s, context 0.4 s, model_review 69.4 s, finding_judge 0.3 s, config: defaults + host + repository (.werkrevisor.toml at main 0323f38). Reviewed head `715f1bab610dcd4d057139a12ec72fe9941c077f` against base `0323f38e4cf90ea3f21c341890cacbb40c4f19dd`, werkrevisor 0.31.5. </details>

ai-reviewer reviewed

2026-10-10 12:31:33 +02:00

ai-reviewer left a comment

werkrevisor: ORANGE / FINDINGS for 715f1bab610dcd4d057139a12ec72fe9941c077f.
The result comment of werkrevisor holds the details.

werkrevisor: ORANGE / FINDINGS for `715f1bab610dcd4d057139a12ec72fe9941c077f`. The result comment of werkrevisor holds the details.

tools/review-rounds Outdated

						
				@@ -0,0 +29,4 @@

				            return json.load(response)

				    for repo in repos:

				        for pr in get(f"{repo}/pulls?state=all&sort=recentupdate&limit=50"):

ai-reviewer commented

2026-10-10 12:31:33 +02:00

API Pagination Limit, severity unknown, counted as medium (werkrevisor)

ORANGE; Jev: real 0.78, severity 2.6 of 3

The script requests limit=50 from the Gitea pulls endpoint and filters by created_at afterwards. If a repository has more than 50 PRs in the requested time range, older PRs are silently dropped. The Werkbaum follow-up node targets "10 to 15 PRs", so the limit is fine in practice, but the script advertises itself as a generic counter and would mislead an operator of an active repo. Either raise the limit, paginate, or document the cap.

**API Pagination Limit**, severity unknown, counted as medium (werkrevisor) ORANGE; Jev: real 0.78, severity 2.6 of 3 The script requests `limit=50` from the Gitea pulls endpoint and filters by `created_at` afterwards. If a repository has more than 50 PRs in the requested time range, older PRs are silently dropped. The Werkbaum follow-up node targets "10 to 15 PRs", so the limit is fine in practice, but the script advertises itself as a generic counter and would mislead an operator of an active repo. Either raise the limit, paginate, or document the cap.

claude-code commented

2026-10-10 12:32:13 +02:00

Fixed in fb3f5b3: the script reads all pages of PRs, 50 per page, until a page is empty.

Fixed in `fb3f5b3`: the script reads all pages of PRs, 50 per page, until a page is empty.

claude-code marked this conversation as resolved

ai-reviewer scheduled this pull request to auto merge when all checks succeed 2026-10-10 12:31:33 +02:00

ai-reviewer requested review from mi 2026-10-10 12:31:33 +02:00

claude-code commented

2026-10-10 12:32:14 +02:00

Review round of 715f1bab: the ORANGE finding, the table row and Scenario#72.02 are fixed in fb3f5b3.

Review round of `715f1bab`: the ORANGE finding, the table row and Scenario#72.02 are fixed in `fb3f5b3`.

mi added 1 commit 2026-10-10 12:32:14 +02:00

review-rounds reads all pages of PRs; clearer table row and Scenario#72.02

werkator build succeeded after 00:27

Details

ai-review/assessment AI review passed; approved by werkrevisor; merge already scheduled; review request of mi withdrawn

Details

ai-review/gate build succeeded after 01:21

Details

werkator/werkrevisor-statistics build succeeded after 01:02

Details

fb3f5b36c3

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

ai-reviewer commented

2026-10-10 12:33:35 +02:00

🟢 werkrevisor: GREEN / PASS

Change kind: mixed.

Reasons

hint 5 changed file(s) that neither weaken the checks nor touch security according to Jev
hint 1 build change(s) of low risk according to Jev: statistics/build.gradle.kts (weakens checks 0.02, risk 0.00, security 0.01)
hint 7 section(s) of the PR document agree with the change summary of openrouter/minimax/minimax-m3 according to Jev

Classifier

7 of 7 units passed all questions (1 PR document(s)).
3 answers ignored because of a confidence below [classifier] min_confidence.

_{Usage: model `openrouter/minimax/minimax-m3`, LLM done, tokens counted before the call 46776, calls 1, tokens in 45202 / out 6885 (reasoning included in out), duration 75.7 s, cost 0.0218 USD (reported), classifier `jev-1.13.0` calls 7, tokens in 5448, cost 0.0002 USD, change summary `openrouter/minimax/minimax-m3` via Minimax calls 1, tokens in 2396 / out 1337, cost 0.0023 USD (reported), PR document check `jev-1.13.0` calls 7, tokens in 10197, cost 0.0004 USD, build file judge `jev-1.13.0` calls 6, tokens in 4015, cost 0.0002 USD, phases: path_judge 0.5 s, classifier 0.5 s, change_summary 16.6 s, doc_check 0.5 s, context 0.4 s, model_review 56.7 s, config: defaults + host + repository (.werkrevisor.toml at main 0323f38).

Reviewed head `fb3f5b36c3d6ade72da348b4f7f0354684be0868` against base `0323f38e4cf90ea3f21c341890cacbb40c4f19dd`, werkrevisor 0.31.5.}

## 🟢 werkrevisor: GREEN / PASS Change kind: `mixed`. ### Reasons - **hint** 5 changed file(s) that neither weaken the checks nor touch security according to Jev - **hint** 1 build change(s) of low risk according to Jev: statistics/build.gradle.kts (weakens checks 0.02, risk 0.00, security 0.01) - **hint** 7 section(s) of the PR document agree with the change summary of openrouter/minimax/minimax-m3 according to Jev ### Classifier 7 of 7 units passed all questions (1 PR document(s)). 3 answers ignored because of a confidence below [classifier] min_confidence. Usage: model `openrouter/minimax/minimax-m3`, LLM done, tokens counted before the call 46776, calls 1, tokens in 45202 / out 6885 (reasoning included in out), duration 75.7 s, cost 0.0218 USD (reported), classifier `jev-1.13.0` calls 7, tokens in 5448, cost 0.0002 USD, change summary `openrouter/minimax/minimax-m3` via Minimax calls 1, tokens in 2396 / out 1337, cost 0.0023 USD (reported), PR document check `jev-1.13.0` calls 7, tokens in 10197, cost 0.0004 USD, build file judge `jev-1.13.0` calls 6, tokens in 4015, cost 0.0002 USD, phases: path_judge 0.5 s, classifier 0.5 s, change_summary 16.6 s, doc_check 0.5 s, context 0.4 s, model_review 56.7 s, config: defaults + host + repository (.werkrevisor.toml at main 0323f38). Reviewed head `fb3f5b36c3d6ade72da348b4f7f0354684be0868` against base `0323f38e4cf90ea3f21c341890cacbb40c4f19dd`, werkrevisor 0.31.5.

ai-reviewer approved these changes

2026-10-10 12:33:35 +02:00

ai-reviewer left a comment

werkrevisor: GREEN / PASS for fb3f5b36c3d6ade72da348b4f7f0354684be0868.
The result comment of werkrevisor holds the details.

werkrevisor: GREEN / PASS for `fb3f5b36c3d6ade72da348b4f7f0354684be0868`. The result comment of werkrevisor holds the details.

ai-reviewer removed review request for mi 2026-10-10 12:33:36 +02:00

ai-reviewer merged commit 5e91298b44 into main

2026-10-10 12:33:40 +02:00

ai-reviewer deleted branch claude-code/pr-doc-hints-as-line-comments

2026-10-10 12:33:40 +02:00

Sign in to join this conversation.

3 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: mi/werkrevisor#72