Agent skill

Screenshot Review

by Inebrio in Inebrio/Routerly

Judge which documentation screenshots a change actually invalidates, on top of the mechanical selector, and get a person's approval before regenerating any of them.

AGPL-3.0Auto-check passed

Install Screenshot Review

skills CLI
$ npx skills add Inebrio/Routerly --skill screenshot-review -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install Inebrio/Routerly screenshot-review --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/Inebrio/Routerly.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/screenshot-review .claude/skills/screenshot-review && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
screenshot-review
GitHub stars
101
Token cost
~769 tokens
SKILL.md length
424 words
Files
1
Skills in repo
17
Repo updated
First seen
Licence
AGPL-3.0

At a glance

Judge which documentation screenshots a change actually invalidates, on top of the mechanical selector, and get a person's approval before regenerating any of them.

  • Works in 7 steps: Run the selector first. → Read the actual diff for that range. git… → Widen when the diff reaches further than… → …
  • Reviewing a diff that touches the dashboard before its screenshots are refreshed
  • Calls npm and git

What it does

Screenshot Review is an agent skill from Inebrio/Routerly. Judge which documentation screenshots a change actually invalidates, on top of the mechanical selector, and get a person's approval before regenerating any of them. Use when reviewing a diff that touches the dashboard before its screenshots are refreshed.

Its SKILL.md is about 770 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: Self-hosted LLM gateway that routes requests across AI providers (OpenAI, Anthropic, Gemini, Mistral, Ollama) using intelligent multi-policy scoring — including an LLM-native… The licence is AGPL-3.0.

When your agent uses it

  • Reviewing a diff that touches the dashboard before its screenshots are refreshed

Example prompts

  • “/screenshot-review”

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. Run the selector first.
  2. Read the actual diff for that range. git diff (or the
  3. Widen when the diff reaches further than the map described. A shared
  4. Show the final list to a person before running anything. State three
  5. Run the capture only after approval, using the frozen composed
  6. Leave the commit to the person. Any resulting change under
  7. Report a failing capture as a failure. If npm run screenshots

What it can do on your machine

Read from SKILL.md and the folder at commit bcf58f1. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • npm
    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use npm and git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Screenshot Review loads about 769 tokens when it runs. Until then it costs about 68 tokens; SKILL.md has 424 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~68
When it runs · the whole SKILL.md, loaded when a task matches
~769

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from Inebrio/Routerly at commit bcf58f1, republished under its AGPL-3.0 licence (© Inebrio). 424 words, ~769 tokens.

Download SKILL.mdSave it as .claude/skills/screenshot-review/SKILL.md (or your agent's skills folder).
name
screenshot-review
description
Judge which documentation screenshots a change actually invalidates, on top of the mechanical selector, and get a person's approval before regenerating any of them. Use when reviewing a diff that touches the dashboard before its screenshots are refreshed.

Screenshot review

The mechanical layer, scripts/screenshots-affected.mjs, is the reproducible base: same diff, same list, every time, no Codex required. This skill is the judgement layer on top of it, for the cases a file-path map cannot decide. It never replaces the selector and it never runs the capture on its own authority.

Steps, in order

  1. Run the selector first.

    npm run screenshots:affected -- --range <range>

    <range> is whatever range is under review (a PR's range, a release branch's range since it diverged, or the default HEAD~1..HEAD). This is the mechanical proposal: one shot name per line, or nothing if the selector found no affected shot. Treat it as the floor of the final list, never as the ceiling.

  2. Read the actual diff for that range. git diff <range> (or the equivalent for the surface under review). Do not judge from file names alone; read what changed.

  3. Widen when the diff reaches further than the map described. A shared component, a layout wrapper, a route change, a style token that several pages consume, anything the mechanical rule in scripts/screenshots/impact.json did not anticipate: add the shots those pages own to the list. Never narrow the list below what the selector already printed. If you have a reason to think one of the selector's own names is not actually affected, say so explicitly in the summary shown to the person in step 4; do not drop it silently.

  4. Show the final list to a person before running anything. State three things: what the selector proposed, what was added and why, and what (if anything) was kept despite looking possibly unnecessary and why it was kept anyway. Wait for that person's approval. This skill never proceeds past this point on its own judgement alone.

  5. Run the capture only after approval, using the frozen composed command with the approved names, comma-separated:

    npm run screenshots -- --only <name>[,<name>...]

    Do not invent flags or behavior beyond this contract and the selector's --range flag; both are frozen elsewhere and this skill only calls them.

  6. Leave the commit to the person. Any resulting change under docs/assets/*.png is committed and pushed by the person, not by this skill. This skill never commits and never pushes on its own authority.

  7. Report a failing capture as a failure. If npm run screenshots exits non-zero, that is the outcome to report, verbatim (the reason it printed on stderr), never rounded up to success and never swallowed. A partial run is not a partial success; the capture script itself writes no partial file on failure, so there is nothing to salvage from it.

© Inebrio, AGPL-3.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/screenshot-review of Inebrio/Routerly.

Open the folder on GitHubat commit bcf58f1

Compare with similar skills

Screenshot Review next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Screenshot Review compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Screenshot Review this skillInebrio/Routerly101—~769Automated safety check: PassAGPL-3.0
Screenshotnexu-io/open-design100k—~291Automated safety check: PassApache-2.0
Tabler Screenshot Makertabler/tabler42k—~1.3kAutomated safety check: PassMIT
OWASP Top 10 Testing with Strixusestrix/strix68k—~1.6kAutomated safety check: PassApache-2.0
Screenshots Marketingnexu-io/open-design100k—~301Automated safety check: PassApache-2.0
PR Screenshotsgithub/awesome-copilot40k—~1.2kAutomated safety check: PassMIT

Similar skills

  • Screenshot

    nexu-io/open-design

    Capture desktop, app windows, or pixel regions across OS platforms.

    100k GitHub stars~291 tokensUpdated yesterday
    Testing & QAAuto-check passed
  • Makes reproducible screenshots of Tabler components and screens for release notes, the website and social posts, using the repository's screenshots app.

    42k GitHub stars~1.3k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Runs Strix's autonomous exploit agents against each OWASP Top 10:2025 category and the API Security Top 10, reporting only what could actually be proven with a proof-of-concept.

    68k GitHub stars~1.6k tokensUpdated today
    SecurityAuto-check passed
  • Screenshots Marketing

    nexu-io/open-design

    Generate marketing screenshots with Playwright. An agent skill from nexu-io/open-design.

    100k GitHub stars~301 tokensUpdated yesterday
    Frontend & DesignAuto-check passed
  • PR Screenshots

    github/awesome-copilot

    Official

    Embed before/after screenshots and annotated images in pull request descriptions.

    40k GitHub stars~1.2k tokensUpdated 2 days ago
    DevelopmentAuto-check passed
  • UI Screenshots

    github/awesome-copilot

    Official

    Capture screenshots of web apps during development using Playwright and PIL.

    40k GitHub stars~1.9k tokensUpdated 2 days ago
    Testing & QAAuto-check passed

More from Inebrio/Routerly

All 17 skills in this repo
  • Docs Conventions

    Inebrio/Routerly

    Where documentation lives in this repository, how a page is structured, and what a change owes the docs.

    101 GitHub stars~624 tokensUpdated 2 days ago
    Auto-check passed
  • Feature Lifecycle

    Inebrio/Routerly

    Run a Tier 1 or Tier 2 feature end to end — feature level, story level, parallelism, integration, and retrospective.

    101 GitHub stars~3k tokensUpdated 2 days ago
    Auto-check passed
  • Story Lifecycle

    Inebrio/Routerly

    Run one user story end to end in its own worktree, from claim to merge.

    101 GitHub stars~905 tokensUpdated 2 days ago
    Auto-check passed
  • Test Conventions

    Inebrio/Routerly

    Where tests live in this repository, how to run them, and what a story owes in coverage.

    101 GitHub stars~467 tokensUpdated 2 days ago
    Auto-check passed
  • Validation Protocol

    Inebrio/Routerly

    How to validate a story against its criteria and write the report, including what counts as evidence and what counts as blocking.

    101 GitHub stars~628 tokensUpdated 2 days ago
    Auto-check passed
  • Analysis Format

    Inebrio/Routerly

    Shape of a feature analysis document (00-analysis.md). An agent skill from Inebrio/Routerly.

    101 GitHub stars~472 tokensUpdated 2 days ago
    Auto-check passed

Questions about Screenshot Review

What does Screenshot Review do?

Judge which documentation screenshots a change actually invalidates, on top of the mechanical selector, and get a person's approval before regenerating any of them. Screenshot Review is an agent skill from Inebrio/Routerly. Judge which documentation screenshots a change actually invalidates, on top of the mechanical selector, and get a person's approval before regenerating any of them.

When should I use Screenshot Review?

Screenshot Review fits situations like: reviewing a diff that touches the dashboard before its screenshots are refreshed.

How do I install Screenshot Review in Claude Code?

Run `npx skills add Inebrio/Routerly --skill screenshot-review -a claude-code`. Or copy the skill folder (.agents/skills/screenshot-review in Inebrio/Routerly) into .claude/skills/screenshot-review in your project. Claude Code loads it when a task matches its description.

How do I install Screenshot Review in Codex?

Run `npx skills add Inebrio/Routerly --skill screenshot-review -a codex`. Or copy the skill folder (.agents/skills/screenshot-review in Inebrio/Routerly) into .agents/skills/screenshot-review in your project. Codex loads it when a task matches its description.

Can I use Screenshot Review in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Inebrio/Routerly --skill screenshot-review -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/screenshot-review, .gemini/skills/screenshot-review, .github/skills/screenshot-review and .opencode/skills/screenshot-review in your project.

What does Screenshot Review need to run?

Going by SKILL.md and its folder, Screenshot Review needs the command-line tools its instructions call (npm and git).

Does Screenshot Review access the network?

SKILL.md contains no URLs. Its commands use npm and git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Screenshot Review safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Screenshot Review use?

Screenshot Review is published under the AGPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Screenshot Review use?

About 769 tokens (SKILL.md is roughly 3.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Screenshot Review?

Skills that share tags, products or a category with Screenshot Review: Screenshot (nexu-io/open-design, 100k stars), Tabler Screenshot Maker (tabler/tabler, 42k stars), OWASP Top 10 Testing with Strix (usestrix/strix, 68k stars) and Screenshots Marketing (nexu-io/open-design, 100k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Screenshot Review?

Inebrio (a GitHub organization) maintains it in Inebrio/Routerly, which has 101 GitHub stars. The repository holds 17 skills in this directory. The repository was last updated on October 8, 2026.

Source: Inebrio/Routerly on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.