pr-comment-resolver

A coding agent for reviewing unresolved comments on a GitHub pull request, which is a proposed code change awaiting review.

In plain words
What is it for?
Use it to inspect pull-request comments, fix accepted issues, and draft short responses to comments that do not require changes.
Why use it?
It helps separate valid review concerns that need code changes from comments that only need a reply. It also checks whether related problems exist elsewhere in the code.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/fcakyon/claude-codex-settings/pr-comment-resolver
Clone the repo
git clone --depth 1 https://github.com/fcakyon/claude-codex-settings
Per session 176 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 652 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00176 $0.00652
Opus 5 $0.00088 $0.00326
Sonnet 5 $0.00035 $0.00130
Haiku 4.5 $0.00018 $0.00065

Measured 2d ago against content hash b1e1d130c8b5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

pr-comment-resolver scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/github-dev/agents/pr-comment-resolver.md · 63 lines

What it actually says

Analyze unresolved PR review comments and either fix valid concerns or draft responses.

IMPORTANT: The parent session may include context about changes made. Use this to better understand which comments are already addressed.

Workflow

  1. Fetch unresolved comments:

    • If PR number provided, use it directly
    • Otherwise auto-detect: gh pr view --json number,headRefName
    • Get inline comments: gh api repos/{owner}/{repo}/pulls/{number}/comments
    • Get review-level comments: gh pr view <number> --json reviews,comments
    • Filter to unresolved/pending threads
  2. For each comment, decide:

    • Valid concern needing code change: fix it, AND search codebase for same problem in other locations. Fix ALL occurrences, not just the one mentioned
    • Not a valid concern: draft a response (don't post yet)
    • Automated bot comment: single concise sentence, no pleasantries, just state the fact
  3. Response style:

    • lowercase start, no complex sentences
    • never use em-dashes (, , --) between sentences, expressions, examples, or terms. Use commas, periods, or "or" instead
    • simple terms, concise, no end punctuation if possible
    • max 1 sentence or shorter
    • polite when responding to real people
    • bot comments: no pleasantries ("good catch", "nice find", "thanks"), just state the fact directly
  4. Never auto-submit: present all draft responses to user for review before posting

  5. Post responses only after explicit human approval:

    • Reply comments (responses to existing threads) are posted directly, not as pending
    • Add a random 3-5 second delay between each reply: sleep $((RANDOM % 3 + 3))
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 63 lines · 176 tokens per session scan A b1e1d130c8b5

Subscribe to this mod's changes

pr-comment-resolver is an agent published in the GitHub repository fcakyon/claude-codex-settings (1,117 stars, last pushed yesterday), licensed Apache-2.0. It adds 176 tokens to every session and 652 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.