Consolidate coderabbitai/claude-plugin into this repo#9
Conversation
Move plugin manifest, subagent, and slash command from coderabbitai/claude-plugin to the root of this repo so a single source of truth hosts both the multi-agent skills and the Claude Code plugin. Skill content (skills/autofix, skills/code-review) is unchanged and already newer than the claude-plugin copies. Next steps (separate PRs): update marketplace.json in anthropics/claude-plugins-official to point at this repo, then archive coderabbitai/claude-plugin.
📝 WalkthroughWalkthroughAdds documentation: a new "Claude Code Plugin" installation section in README, plus two new docs— Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Poem
🚥 Pre-merge checks | ✅ 4✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
✨ Simplify code
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
commands/review.md (1)
54-65: Make the runnable command unambiguous for default type handling.Line 55 shows
-t <type>while Lines 58–64 describe defaults/optional--base. Consider showing the exact interpolated form so users don’t copy a placeholder literally.Suggested doc tweak
-```bash -coderabbit review --plain -t <type> -``` +```bash +# If no type is provided, default to "all" +coderabbit review --plain -t ${type:-all} ${base:+--base "$base"} +```🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@commands/review.md` around lines 54 - 65, Update the example invocation "coderabbit review --plain -t <type>" to an unambiguous, copy-pastable form that shows the default behavior and optional base flag; specifically document the command using a shell-interpolated form that defaults type to "all" and only includes --base when provided (e.g., show the invocation pattern that uses shell parameter expansion for type default and conditional base inclusion), and mention the meaning of the token "type" and the optional "--base" so users won't paste the placeholder literally; update the block containing "coderabbit review --plain -t <type>" and the surrounding explanation to reference these exact tokens ("type", "--base") and the default "all".
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@README.md`:
- Around line 34-37: The README currently labels the two lines "/plugin
marketplace update" and "/plugin install coderabbit" as bash, which is
misleading because they are Claude interactive slash commands; change the fenced
code block language from bash to plain text (or annotate as "Claude slash
commands") and add a short note or caption indicating these are interactive
slash commands to be entered in Claude, not shell commands, referencing the
exact lines containing "/plugin marketplace update" and "/plugin install
coderabbit".
---
Nitpick comments:
In `@commands/review.md`:
- Around line 54-65: Update the example invocation "coderabbit review --plain -t
<type>" to an unambiguous, copy-pastable form that shows the default behavior
and optional base flag; specifically document the command using a
shell-interpolated form that defaults type to "all" and only includes --base
when provided (e.g., show the invocation pattern that uses shell parameter
expansion for type default and conditional base inclusion), and mention the
meaning of the token "type" and the optional "--base" so users won't paste the
placeholder literally; update the block containing "coderabbit review --plain -t
<type>" and the surrounding explanation to reference these exact tokens ("type",
"--base") and the default "all".
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Central YAML (base), Organization UI (inherited)
Review profile: CHILL
Plan: Pro Plus
Run ID: 13fc508d-50ce-4d96-a596-338e223ee1bd
⛔ Files ignored due to path filters (1)
.claude-plugin/plugin.jsonis excluded by!**/*.json
📒 Files selected for processing (3)
README.mdagents/code-reviewer.mdcommands/review.md
📜 Review details
🧰 Additional context used
🪛 LanguageTool
agents/code-reviewer.md
[grammar] ~23-~23: Use a hyphen to join words.
Context: ...nd potential bugs, edge cases, and error handling issues ## When to Use Use thi...
(QB_NEW_EN_HYPHEN)
🔇 Additional comments (3)
agents/code-reviewer.md (1)
43-68: Workflow structure is clear and actionable.The staged flow (context → run review → analyze → recommend → interactive resolution) is well organized and easy to follow.
README.md (1)
30-37: Nice addition: plugin install path improves onboarding.This section makes the Claude Code path discoverable without changing existing
npx skills addflow.commands/review.md (1)
4-17: No issue here. Shell built-ins and read-only utilities likepwd,head, andgrepare automatically allowed by Claude Code without requiring explicit listing inallowed-tools. Theallowed-toolsfrontmatter correctly specifies access to coderabbit, cr, and git tools; the context block commands will execute without restriction.> Likely an incorrect or invalid review comment.
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
agents/code-reviewer.md (1)
50-53: Use the canonical CLI invocation pattern documented for the review command.
coderabbit review --plainhere is underspecified compared tocommands/review.md:48-73, which standardizescoderabbit review --plain -t "${type:-all}" ${base:+--base "$base"}. Recommend mirroring that pattern (or explicitly linking to it) to prevent argument drift between docs.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@agents/code-reviewer.md` around lines 50 - 53, Documentation uses an underspecified CLI invocation; replace the example `coderabbit review --plain` with the canonical pattern `coderabbit review --plain -t "${type:-all}" ${base:+--base "$base"}` (or add an explicit link/reference to the canonical pattern in commands/review.md) so callers pass the standardized type and optional base arguments; update the text around the example in the "Run CodeRabbit Review" section and ensure any copy of the command mirrors the `-t "${type:-all}" ${base:+--base "$base"}` template.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@agents/code-reviewer.md`:
- Around line 69-97: The review taxonomy in the "## Review Categories" block of
agents/code-reviewer.md uses four tiers (Critical, High Priority, Medium
Priority, Low Priority) which conflicts with the three-group model used by
commands/review.md; update agents/code-reviewer.md so its headings and item
mappings match commands/review.md's groups (Critical, Suggestions, Positive) —
i.e., collapse or remap High Priority/Medium Priority/Low Priority entries into
Suggestions or Positive as appropriate, keep the Critical section intact, and
ensure any descriptive bullets under the existing headings are moved/renamed
accordingly; verify any references to "## Review Categories" or the old tier
names in this file are updated to the new names to keep agent and slash-command
behavior consistent.
---
Nitpick comments:
In `@agents/code-reviewer.md`:
- Around line 50-53: Documentation uses an underspecified CLI invocation;
replace the example `coderabbit review --plain` with the canonical pattern
`coderabbit review --plain -t "${type:-all}" ${base:+--base "$base"}` (or add an
explicit link/reference to the canonical pattern in commands/review.md) so
callers pass the standardized type and optional base arguments; update the text
around the example in the "Run CodeRabbit Review" section and ensure any copy of
the command mirrors the `-t "${type:-all}" ${base:+--base "$base"}` template.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Central YAML (base), Organization UI (inherited)
Review profile: CHILL
Plan: Pro Plus
Run ID: 50aaf92c-4d41-4e59-8d44-ae9d18ada8c2
⛔ Files ignored due to path filters (1)
.claude-plugin/plugin.jsonis excluded by!**/*.json
📒 Files selected for processing (3)
README.mdagents/code-reviewer.mdcommands/review.md
✅ Files skipped from review due to trivial changes (2)
- README.md
- commands/review.md
📜 Review details
🧰 Additional context used
🪛 LanguageTool
agents/code-reviewer.md
[grammar] ~23-~23: Use a hyphen to join words.
Context: ...nd potential bugs, edge cases, and error handling issues ## When to Use Use thi...
(QB_NEW_EN_HYPHEN)
1 participant
Summary
.claude-plugin/plugin.json,agents/code-reviewer.md, andcommands/review.mdat the repo root, making the whole repo installable as the officialcoderabbitClaude Code plugin.README.mdwith a Claude Code plugin install section alongsidenpx skills add.skills/,agents/,commands/at root) scales to future agents that support subagents, slash commands, hooks, etc.Skill content (
skills/autofix,skills/code-review) is unchanged — the versions here are already newer than the copies in coderabbitai/claude-plugin, so nothing to merge.Follow-ups (separate PRs, in order)
coderabbitentry in.claude-plugin/marketplace.jsonto point athttps://github.com/coderabbitai/skills.git.apps/cliinstaller to projectagents/andcommands/into per-agent paths where supported.Test plan
/plugin installagainst it in Claude Code to confirm the plugin loads (agents + slash command + skills all resolve).npx skills add coderabbitai/skills#claude_plugin_shiftto confirm the multi-agent installer still works unchanged./coderabbit:reviewslash command is discoverable after install.Summary by CodeRabbit
Documentation