Agent skill
inline-pr-comments
Post inline PR review comments on specific lines. Use this skill to deliver code review feedback as inline comments rather than a single summary.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/inline-pr-comments
SKILL.md
Inline PR Comments Skill
Use this skill to post code review feedback as inline comments on specific lines in a PR.
When to Use
- After completing a code review (use with /claude-review, /claude-security-review, /claude-tob-review)
- When you have specific line-by-line feedback to deliver
- To make review feedback more actionable
Instructions
Use MCP inline comment tool, not gh api:
- For each issue on a changed line, call:
mcp__github_inline_comment__create_inline_comment
- Required args:
pathbodyline(single-line), orstartLine+line(multi-line)side: "RIGHT"unless intentionally commenting old code
- After posting inline comments, update the Claude sticky summary comment with:
mcp__github_comment__update_claude_comment
- Do not use
gh apihere; action tool permissions often block non-git Bash commands.
Comment Fields (create_inline_comment)
path- File path relative to repo rootline- Line number in the NEW version of the file (right side of diff)startLine+line- For comments spanning multiple linesside-RIGHT(new code) orLEFT(old code), default toRIGHTbody- Markdown-formatted feedback
Limitations
- Can only comment on lines that appear in the diff (changed/added lines)
- Comments on unchanged lines will fail with "Line could not be resolved"
Handling Non-Diff Findings
When you discover issues in code NOT changed by the PR:
- Include in summary body - Always report in the
"body"field - Format clearly - Use a dedicated section "## Observations Outside This PR"
- Be actionable - Include file:line references so author can follow up
- Don't block - These are informational; don't use
REQUEST_CHANGESfor non-diff issues
Example: update the sticky comment to include non-diff findings:
tool: mcp__github_comment__update_claude_comment
args:
body: |
## Review Summary
[inline feedback summary]
## Observations Outside This PR
While reviewing, I noticed:
- `src/utils/foo.ts:142`: Pre-existing null check missing
- `src/core/bar.ts:78-82`: Similar pattern to line 45 issue - consider deduping
Feedback Guidelines
| Feedback Type | In Diff? | Where to Put It |
|---|---|---|
| Specific code issue | ✅ Yes | Inline comment on that line |
| Pattern repeated across files | ✅ Yes | Inline on first occurrence + note "same issue in X, Y, Z" |
| Related issue found | ❌ No | Summary body under "Observations Outside This PR" |
| Pre-existing bug discovered | ❌ No | Summary body (consider separate issue if critical) |
| Overall architecture concern | N/A | Summary body |
| Approve PR | N/A | Out of scope for this skill; keep feedback in sticky comment |
| Changes requested | N/A | Sticky comment; never use REQUEST_CHANGES |
Be concise. Group minor style issues together.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?