Tatsh / wiswa / 34448743958

100%

push

github

project: add prose rules and a prose-lint skill

The generator wrote no prose rule set. Style guidance for comments,
docstrings, and Markdown lived inside the copy-editor agent alone,
invisible to a direct edit or to another agent.

The new rule file bans contractions, en and em dashes, seventeen verbs,
a list of idioms, several sentence patterns, colon labels, and four
heading forms. It applies to Markdown, plain comments, formal
documentation comments such as Numpydoc and Doxygen, user-facing
strings, and commit messages. Acronyms take full uppercase, and an
entity such as Ofcom follows common usage instead.

The prose-lint skill audits against the rule file with a set of greps
and applies the fixes. The copy-editor agent cites the same file in
three places and overrides its softer defaults. A borderline violation
is rewritten rather than excused. The agent also drops the upper-lower
acronym rule. The new file reverses it.

Both new files render from templates with no Jinja conditionals. The
generated copy under .claude matches byte for byte.

Signed-off-by: Andrew Udvare <audvare@gmail.com>

788 of 788 branches covered (100.0%)

Branch coverage included in aggregate %.

2040 of 2040 relevant lines covered (100.0%)

5.0 hits per line