Post review

Skill rperez030/accessible-agent-skills/skills/post-review

Claude Code skills and hooks built to make working with Claude more accessible for developers who use assistive technology.

Install

npx -y skills add rperez030/accessible-agent-skills --skill post-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Post a collected GitHub PR review with inline comments as a single review submission. Use when the user is ready to post a PR review after all feedback has been gathered.

SKILL.md

2.5 KB, 543 tokens by cl100k_base, as published. Nobody here has run it

Post PR Review

This skill posts a GitHub PR review. All feedback must already be collected and agreed upon before invoking this skill.

Rules — follow these exactly, no exceptions

Never post inline comments before the review is ready. This skill is the signal that the review is ready to post. Do not post anything until the user invokes it.
Ask whether to include a summary comment (body). If the user wants one, draft it and present for explicit approval before posting. Do not auto-generate a summary. The summary is posted under the user's name and must reflect their voice and accurately represent the severity of findings. If no summary is needed, use an empty string for the body field.
Always post everything as a single API call using gh api repos/{owner}/{repo}/pulls/{number}/reviews with a JSON body that includes the summary body, the event field, and all inline comments in one array. Never post inline comments separately via /pulls/{number}/comments — that creates multiple disconnected review entries.
Use --input - with a heredoc to pass the full JSON payload, not individual --field arguments. This is the only reliable way to include multiple inline comments in a single call.
Verify the result by checking that only one new review ID appears and that its state matches the intended event (e.g. CHANGES_REQUESTED).

Template

gh api repos/{owner}/{repo}/pulls/{number}/reviews \
  --method POST \
  --input - <<'EOF'
{
  "commit_id": "{full_sha}",
  "event": "REQUEST_CHANGES",
  "body": "{summary comment}",
  "comments": [
    {
      "path": "{file path}",
      "line": {line number},
      "side": "RIGHT",
      "body": "{inline comment text}"
    }
  ]
}
EOF

Valid values for event: REQUEST_CHANGES, APPROVE, COMMENT.

If something goes wrong

If comments landed as separate reviews by mistake:

Delete each inline comment via DELETE /repos/{owner}/{repo}/pulls/comments/{comment_id}
Dismiss the bad review via PUT /repos/{owner}/{repo}/pulls/{number}/reviews/{review_id}/dismissals
Note: GitHub does not allow deleting submitted reviews, only dismissing them — dismissed reviews remain visible in the PR timeline
Repost using the template above

Gives 0 of the 12 instructions most review quality skills give in 543 tokens

Counted across 1,048 of the 1,783 authors here whose files we hold, read 2026-08-06

ask questions one at a timein 82 of 1048, across 54 files
provide a recommended answer for each questionin 73 of 1048, across 45 files
explore the codebase instead of asking answerable questionsin 66 of 1048, across 37 files
resolve dependencies between decisions one-by-onein 42 of 1048, across 15 files
interview the user relentlessly about the planin 39 of 1048, across 12 files
order findings by severityin 29 of 1048
resolve each branch of the decision treein 28 of 1048, across 5 files
run a grilling sessionin 26 of 1048, across 5 files
update CONTEXT.md immediately when a term is resolvedin 26 of 1048, across 9 files
propose precise canonical terms for vague languagein 25 of 1048, across 6 files
create documentation files lazilyin 24 of 1048, across 5 files
use the domain-modeling skillin 22 of 1048, across 3 files

Said here and by no other author read

ask to include a summary comment before posting
present the summary for explicit approval before posting
post all comments as a single API call
pass the full JSON payload using a heredoc
verify one new review ID appears with the intended state
delete accidentally separate inline comments

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.