skills.rest
chughtapan
chughtapanCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

dogfood

Evaluate GitHub issues, PRs, or Markdown files with a four-axis rubric and publish a verdict.

Updated Apr 18, 2026
Search Tags:#readability#GitHub#rubric#artifact#friction#verdict#dogfood
One-click install
npx skills add https://github.com/chughtapan/safer-by-default --skill dogfood-chughtapan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: dogfood
Source: https://github.com/chughtapan/safer-by-default/tree/main/skills/dogfood
Command: npx skills add https://github.com/chughtapan/safer-by-default --skill dogfood-chughtapan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Read published artifacts (spec, design doc, PR body, issue comment, investigation writeups) with NO prior session context and report on how a cold-start reader would feel picking it up. It scores four numeric axes (clarity, completeness, actionability, trust) plus a friction list with evidence, emits a SHIP / REVISE / REJECT verdict, and publishes the report back to the artifact's thread. Use when an artifact is about to be handed off and you want portability checks before the next modality runs. Do NOT use to rewrite the artifact; this is a reading modality, not a writing one.

Core Features & Use Cases

  • Resolve an artifact payload (title, body, labels) into a self-contained read.
  • Evaluate with a rigorous rubric (4 axes) and friction log, then publish a verdict to the artifact's GitHub thread.
  • Publish a portable dogfood report suitable for upstream authors and reviewers to inform revisions.

Quick Start

Provide a single artifact reference (GitHub issue/PR or local file) to run a cold-start dogfood evaluation and receive a structured report.

Frequently Asked Questions about dogfood

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a cold-start review on a GitHub PR or issue without prior context?▼

To perform a cold-start review, provide a single artifact reference, such as a GitHub issue, PR, or local Markdown file. The Skill dispatches a subagent to evaluate the payload with no prior session context and returns a structured dogfood report.

What is a dogfood report for a design doc or specification?▼

A dogfood report is a structured evaluation of a published artifact, assessing how a new reader perceives it. It scores clarity, completeness, actionability, and trust, then emits a SHIP, REVISE, or REJECT verdict with a friction list.

Can I use this to rewrite my spec or design doc based on the review?▼

No, this is a reading modality, not a writing one. The Skill evaluates artifacts to provide portability checks and structured verdicts before handoff, but it does not rewrite or modify the original artifact.

Does the evaluation support local Markdown files or only GitHub threads?▼

The evaluation supports both GitHub issues and PRs as well as local Markdown files. It resolves the artifact payload, applies a four-axis rubric, and can publish the resulting verdict directly back to the artifact's GitHub thread.

What is the best way to check if an artifact is ready for handoff to reviewers?▼

Use this Skill to run a cold-start dogfood evaluation before handoff. It applies a rigorous four-axis rubric, collects evidence, and generates a portable report with a SHIP, REVISE, or REJECT verdict to inform necessary revisions.

When should I not use an automated artifact review for my pull requests?▼

You should avoid using this Skill if your goal is to rewrite or directly edit the artifact content. It is designed strictly for reading and evaluating artifacts to produce structured friction logs and verdicts, not for modifying the original text.

Related Skills

Explore More Software Engineering Skills→
chughtapanchughtapan

safer-docs-reader

Aggregate four haiku sub-agent verdicts into a severity-weighted SHIP or REVISE decision.

Community
Advanced
Yeachan-HeoYeachan-Heo
32.6k

review

Evaluate plan files in .omx/plans/ with Critic-based review verdicts.

Community
Basic
richkuorichkuo
45

fableplan-work-on-issue

Plan and implement GitHub issues with Fable 5 subagents and create pull requests.

Community
Advanced
enigmaicon-engenigmaicon-eng
1

advanced-evaluation

Score or compare LLM outputs with rubric-based LLM-as-judge evaluation and structured JSON results.

Community
Advanced