werkrevisor 0.9.0: deterministic writing rules, jev-eval with config layers #6

Merged

mi merged 7 commits from claude-code/deterministic-writing-rules into main 2026-10-05 16:37:00 +02:00

Conversation 4 Commits 7 Files Changed 19

+912 -70

claude-code commented 2026-10-05 16:18:29 +02:00

Collaborator

Copy Link

Copy Source

werkrevisor 0.9.0: deterministic writing rules for changed docs, and jev-eval with the config layers of the review.

• Changed Markdown lines are checked against the AGENTS.md writing style: one sentence per line, line length, sentences per paragraph. Violations give ORANGE; the model still runs.
• The unreliable Jev question about the writing rules is removed.
• tools/jev-eval reads defaults < host < repository config.

Details: the PR document under docs/prs/.

🤖 Generated with Claude Code

werkrevisor 0.9.0: deterministic writing rules for changed docs, and jev-eval with the config layers of the review. - Changed Markdown lines are checked against the AGENTS.md writing style: one sentence per line, line length, sentences per paragraph. Violations give ORANGE; the model still runs. - The unreliable Jev question about the writing rules is removed. - `tools/jev-eval` reads defaults < host < repository config. Details: the PR document under `docs/prs/`. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

claude-code added 6 commits 2026-10-05 16:18:30 +02:00

docs: PR#6 document for deterministic writing rules and jev-eval config layers ... db118ae6d1

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

Deterministic writing rules for changed docs: one sentence per line, line length, paragraph length ... 4f3c9c5c9d

Only added or changed lines of Markdown docs and PR documents count.
A violation gives ORANGE with one aggregated reason; the model still reviews the PR.
The new section [writing] is required and may be set by the repository config.

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

Remove the Jev question about the writing rules ... 118c9926dd

In the jev-eval measurement it flagged all 34 sections, so it did not tell good from bad text.
The countable writing rules are now checked deterministically.

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

jev-eval uses the config layers of the review ... 860f8d102a

--repo-config puts a repository config on top of the host config, with the same rules as the review;
the default is .werkrevisor.toml in the current directory, --no-repo-config uses the host config alone.

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

werkrevisor 0.9.0: README and operations note for [writing] and jev-eval --repo-config ... 8c8f7a9124

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

docs: complete the PR#6 document with the solution and the answered questions ... 678985fefa

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

claude-code marked the pull request as ready for review 2026-10-05 16:18:39 +02:00

ai-reviewer commented 2026-10-05 16:18:45 +02:00

Collaborator

Copy Link

Copy Source

🔴 werkrevisor: RED / CHANGES_REQUIRED

Change kind: mixed.

Reasons

• YELLOW test changes are not purely additive: tests/test_classifier.py:197 removes a test or assertion
• YELLOW change kind 'mixed' (main code, tests, other files, docs); AI-approvable are only: docs only, tests only added, docs and tests, refactoring with unchanged tests
• YELLOW risky paths changed: .werkrevisor.toml (built in), config/hsadmin-ng.toml (config/), config/werkprobe.toml (config/) and 1 more
• RED the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR

_{Usage: model `openrouter/anthropic/claude-sonnet-5.5`, LLM skipped (invalid .werkrevisor.toml), duration 0.2 s, no model call, cost 0 USD, config: defaults + host + repository (.werkrevisor.toml at base 667cbe9).
Reviewed head `678985fefabf879c43818fb6ebb4ff91c6cd6e55` against base `667cbe94bec1f7bd82379f1cf2d8f376a3b50412`, werkrevisor 0.8.0.}

<!-- werkrevisor:678985fefabf879c43818fb6ebb4ff91c6cd6e55 --> <!-- werkrevisor-verdict:RED --> ## 🔴 werkrevisor: RED / CHANGES_REQUIRED Change kind: `mixed`. ### Reasons - **YELLOW** test changes are not purely additive: tests/test_classifier.py:197 removes a test or assertion - **YELLOW** change kind 'mixed' (main code, tests, other files, docs); AI-approvable are only: docs only, tests only added, docs and tests, refactoring with unchanged tests - **YELLOW** risky paths changed: .werkrevisor.toml (built in), config/hsadmin-ng.toml (config/**), config/werkprobe.toml (config/**) and 1 more - **RED** the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR _{Usage: model `openrouter/anthropic/claude-sonnet-5.5`, LLM skipped (invalid .werkrevisor.toml), duration 0.2 s, no model call, cost 0 USD, config: defaults + host + repository (.werkrevisor.toml at base 667cbe9).<br> Reviewed head `678985fefabf879c43818fb6ebb4ff91c6cd6e55` against base `667cbe94bec1f7bd82379f1cf2d8f376a3b50412`, werkrevisor 0.8.0.}

ai-reviewer requested changes 2026-10-05 16:18:45 +02:00

Dismissed

ai-reviewer left a comment

Collaborator

Copy Link

Copy Source

werkrevisor: RED / CHANGES_REQUIRED for 678985fefabf879c43818fb6ebb4ff91c6cd6e55.

• the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR
  Details: #6 (comment)

werkrevisor: RED / CHANGES_REQUIRED for `678985fefabf879c43818fb6ebb4ff91c6cd6e55`. - the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR Details: http://git.javagil.de/mi/werkrevisor/pulls/6#issuecomment-834 <!-- werkrevisor-review:678985fefabf879c43818fb6ebb4ff91c6cd6e55 -->

claude-code added 1 commit 2026-10-05 16:30:40 +02:00

A question mark ends a sentence; an answered question puts its answer on a continuation line ... 3da203c510

The open-questions rule already counts the whole list item as checked, not its source lines.

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

ai-reviewer commented 2026-10-05 16:30:40 +02:00

Collaborator

Copy Link

Copy Source

🔴 werkrevisor: RED / CHANGES_REQUIRED

Change kind: mixed.

Reasons

• YELLOW test changes are not purely additive: tests/test_classifier.py:197 removes a test or assertion
• YELLOW change kind 'mixed' (main code, tests, other files, docs); AI-approvable are only: docs only, tests only added, docs and tests, refactoring with unchanged tests
• YELLOW risky paths changed: .werkrevisor.toml (built in), config/hsadmin-ng.toml (config/), config/werkprobe.toml (config/) and 1 more
• RED the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR

_{Usage: model `openrouter/anthropic/claude-sonnet-5.5`, LLM skipped (invalid .werkrevisor.toml), duration 0.2 s, no model call, cost 0 USD, config: defaults + host + repository (.werkrevisor.toml at base 667cbe9).
Reviewed head `3da203c5106ad337ef4384076b69f1387e8df5ce` against base `667cbe94bec1f7bd82379f1cf2d8f376a3b50412`, werkrevisor 0.8.0.}

<!-- werkrevisor:3da203c5106ad337ef4384076b69f1387e8df5ce --> <!-- werkrevisor-verdict:RED --> ## 🔴 werkrevisor: RED / CHANGES_REQUIRED Change kind: `mixed`. ### Reasons - **YELLOW** test changes are not purely additive: tests/test_classifier.py:197 removes a test or assertion - **YELLOW** change kind 'mixed' (main code, tests, other files, docs); AI-approvable are only: docs only, tests only added, docs and tests, refactoring with unchanged tests - **YELLOW** risky paths changed: .werkrevisor.toml (built in), config/hsadmin-ng.toml (config/**), config/werkprobe.toml (config/**) and 1 more - **RED** the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR _{Usage: model `openrouter/anthropic/claude-sonnet-5.5`, LLM skipped (invalid .werkrevisor.toml), duration 0.2 s, no model call, cost 0 USD, config: defaults + host + repository (.werkrevisor.toml at base 667cbe9).<br> Reviewed head `3da203c5106ad337ef4384076b69f1387e8df5ce` against base `667cbe94bec1f7bd82379f1cf2d8f376a3b50412`, werkrevisor 0.8.0.}

ai-reviewer requested changes 2026-10-05 16:30:41 +02:00

ai-reviewer left a comment

Collaborator

Copy Link

Copy Source

werkrevisor: RED / CHANGES_REQUIRED for 3da203c5106ad337ef4384076b69f1387e8df5ce.

• the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR
  Details: #6 (comment)

werkrevisor: RED / CHANGES_REQUIRED for `3da203c5106ad337ef4384076b69f1387e8df5ce`. - the changed .werkrevisor.toml is invalid: .werkrevisor.toml: unknown section [writing]; after the merge every review would end with ERROR Details: http://git.javagil.de/mi/werkrevisor/pulls/6#issuecomment-837 <!-- werkrevisor-review:3da203c5106ad337ef4384076b69f1387e8df5ce -->

mi merged commit 2c41b912f4 into main 2026-10-05 16:37:00 +02:00

mi deleted branch claude-code/deterministic-writing-rules 2026-10-05 16:37:00 +02:00

Sign in to join this conversation.

Reviewers

No Reviewers

ai-reviewer

Labels

No items

No labels

Milestone

No items

No Milestone

Projects

Clear projects

No projects

Assignees

Clear assignees

ai-reviewer claude-code mi

No Assignees

2 Participants

Notifications

Subscribe

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: mi/werkrevisor#6