feat(writing-style): add quoting exemption, habits, and delivery scan - #1
Merged
Conversation
The skill gained three sections. A quoting exemption keeps tool output, stack traces, and diffs verbatim, so reports stop corrupting their own evidence. A habits list names the tells of machine-written text. A pre-delivery scan checks the mechanical rules before output leaves a command. The work context gained rules for leading with the answer, scaling headings to the artifact, and matching the target repo's spelling. The spelling rule detects the repo's convention rather than hardcoding one, which keeps forge repo-agnostic. CLAUDE.md and AGENTS.md described a testing loop that does not match how forge installs. Claude Code reads a per-version marketplace cache, never this checkout, so editing a file here changes nothing until the change ships. The ignore rule now covers machine-specific agent rules under .agents/rules/.
Adds writing-style rules that change what every forge command produces, so this is a MINOR bump. VERSION.md and plugin.json move together. The plugin.json field gates whether claude plugin update sees the release.
There was a problem hiding this comment.
Pull request overview
This PR expands the writing-style reference skill with a quoting exemption, a machine-writing “habits to avoid” list, and a pre-delivery mechanical scan. It also updates docs to reflect the actual forge installation/testing loop and bumps the plugin to v0.6.0.
Changes:
- Add quoting boundaries, habit anti-patterns, and a “Before delivering” scan to
skills/writing-style. - Clarify docs (
CLAUDE.md,AGENTS.md) around release/update vs local-path iteration. - Cut the v0.6.0 release across
VERSION.md,.claude-plugin/plugin.json, andCHANGELOG.md.
Reviewed changes
Copilot reviewed 8 out of 9 changed files in this pull request and generated no comments.
Show a summary per file
| File | Description |
|---|---|
| VERSION.md | Bumps the repo version marker to 0.6.0. |
| .claude-plugin/plugin.json | Bumps plugin version to 0.6.0 for distribution. |
| CHANGELOG.md | Adds v0.6.0 release notes describing the style-skill and doc updates. |
| skills/writing-style/SKILL.md | Adds quoting exemption, habit list, and pre-delivery scan; refines work/creative context rules. |
| skills/infrastructure-as-code/SKILL.md | Rewords the skill description into shorter sentences without changing keywords. |
| README.md | Updates the writing-style skill summary to mention habits + pre-delivery scan. |
| CLAUDE.md | Updates testing instructions to match marketplace cache + release/update workflow. |
| AGENTS.md | Mirrors the updated testing instructions for non-Claude Code agents. |
| .gitignore | Adds ignore entry intended for machine-specific agent rules. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
The
writing-styleskill gains three sections. A quoting exemption keeps tool output, stack traces, diffs, and cited text verbatim. A habits list names the tells of machine-written text, such as "not X, but Y" framing and inflated stakes. A pre-delivery scan checks the mechanical rules before output leaves a command. The work context adds rules for leading with the answer, scaling headings to the artifact, and matching the target repo's spelling. The branch also cuts the v0.6.0 release soclaude plugin updatecan see the change.Why
Reports were restyling quoted tool output to fit the prose rules, which corrupted their own evidence. The style rules needed an explicit boundary between prose forge writes and material it reproduces. The habits list and delivery scan close the gap between having rules and applying them. Separately,
CLAUDE.mdandAGENTS.mddescribed a testing loop that does not match how forge installs. Claude Code reads a per-version marketplace cache, never this checkout. The docs now describe the release-then-update loop and the local-path shortcut.How
Every change is a prompt file, so there is no code to build. The spelling rule detects the target repo's convention instead of hardcoding American English, which keeps forge repo-agnostic. The delivery scan excludes flag names and number ranges from the dash search to avoid false hits. The habits yield to established voice in the creative context, so existing manuscripts keep their style.
.gitignorenow covers machine-specific agent rules under.agents/rules/.Testing
This repo has no automated test suite. The plugin is prompt files, and testing means invoking the commands against real scenarios. The new sections were reviewed against the skill's existing structure for contradictions, and none were found. Behavioral verification happens after release by running
claude plugin update forge@forge-marketplaceand exercising a command that writes prose.🤖 Generated with Claude Code