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.
npx agentmods add skills/ruddercode/rudder/address-pr-commentsnpx skills add RudderCode/Rudder --skill address-pr-commentsgit clone --depth 1 https://github.com/RudderCode/RudderWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00091 | $0.00950 |
| Opus 5 | $0.00046 | $0.00475 |
| Sonnet 5 | $0.00018 | $0.00190 |
| Haiku 4.5 | $0.00009 | $0.00095 |
Grade A, and why
address-pr-comments 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.
How it starts
The opening of the file, as written. The whole thing — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Address PR Comments
Pull review comments off the current branch's GitHub PR, dedupe them against the current HEAD, and apply or decline each one with a written reason.
Workflow
- Locate the PR for the current branch:
gh pr view --json number,url,headRefName,state
If no PR exists for the current branch, report no PR found for current branch and stop.
- Fetch both top-level (issue) comments and inline review comments:
gh api repos/{owner}/{repo}/issues/{number}/comments --paginate
gh api repos/{owner}/{repo}/pulls/{number}/comments --paginate
{owner}/{repo} is available from gh repo view --json nameWithOwner or from git remote get-url origin.
{number} comes from step 1.
Both endpoints are required — Greptile's top-level summary lives on issues/{number}/comments, while its inline findings live on pulls/{number}/comments.
- Group and de-duplicate:
- Dedupe by
(path, line, author, body-hash)— re-runs of the same finding share a body so they collapse, but distinct findings on the same line (e.g. a P1 and a P2 on the same line) have different bodies and both survive. When a body collides, keep only the most recent. - Discard comments authored by the current user only if they are not feedback to act on (the user's own clarifying replies or self-reviews are not directives; keep them if they contain an explicit instruction or correction directed at the AI).
- Discard comments from deploy bots:
vercel,github-actions, etc. - If a comment thread is resolved on GitHub, skip it unless the user explicitly asks otherwise.
-
Read the code at each cited location. Open the file at
path:lineand confirm the issue is still present at the currentHEAD. A previous commit on this branch may have already fixed it. -
For each unresolved comment, choose one of three actions and record which:
- Fix: apply the smallest correct change. If the comment suggests a patch and you agree with it, follow it; if a smaller or different fix is more correct, do that and explain why in your report.
- Decline: if the comment is wrong, doesn't apply at the current
HEAD, or is purely informational ("note, not a defect"), skip the change and write a one-line reason. - Defer to user: if the comment requires a judgement call, design decision, or context you don't have, do not guess — surface it in the report and stop on that comment.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago First seen · 71 lines · 91 tokens per session scan A a518815b996e
address-pr-comments is a skill published in the GitHub repository RudderCode/Rudder (23 stars, last pushed 14d ago), licensed Apache-2.0. It adds 91 tokens to every session and 950 once invoked, about $0.0005 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.
Other skills, from other repositories
cuopt-numerical-optimization-formulation
LP, MILP, QP — concepts, problem-text parsing, and formulation patterns (parameters, constraints, decisions, objective). Concepts only; no API.
deepstream-sop
Use this skill when building, deploying, evaluating, debugging, or measuring latency for the DeepStream SOP Inference Microservice — a GPU-accelerated FastAPI service that detects whether operators perform assembly-line steps in order via event boundary detection (GEBD) plus VLM classification. Trigger even if the…
digital-health-clinical-asr-eval
Stage 3 of Clinical ASR Flywheel. Score a NeMo manifest, produce the five-section KER leaderboard (by-ipasource diagnostic). Not for ASR auth (/riva-asr).
digital-health-clinical-asr-finetune
Stage 4 of the Clinical ASR Flywheel. Use when priority KER is above 0.3 to run stock NeMo SFT on Parakeet TDT v2 and offline cycle N+1 re-eval. NOT for generic word boosting (use /finetune-asr).
aiq-research
Use when asked to run deep research or AI-Q research through a reachable NVIDIA AI-Q Blueprint backend.
amc-run-rtsp-calibration
Calibrate a new dataset from live RTSP camera streams via the AutoMagicCalib REST API. Use when the user provides RTSP URLs or asks to calibrate live cameras; VIOS records clips, AMC ingests them, then runs calibration.