claude-skills / screens
CLAUDE.md0 starsChanged 30 days agoMITView on GitHub
Rules and commands bound to /screens's own tooling. Cross-cutting rules (run-log.sh, lib-orchestrator.sh) live in the root CLAUDE.md.
In this repository13 other files it gives its agents.
- audit/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- delegate/CLAUDE.mdCLAUDE.md
- plan-it/CLAUDE.mdCLAUDE.md
- ship/CLAUDE.mdCLAUDE.md
- store-assets/CLAUDE.mdCLAUDE.md
- auditSkill
- delegateSkill
- plan-itSkill
- readinessSkill
- screensSkill
- shipSkill
- store-assetsSkill
The file
Files
/screens internals
Rules and commands bound to /screens's own tooling. Cross-cutting rules (run-log.sh,
lib-orchestrator.sh) live in the root CLAUDE.md.
Commands
| Command | Purpose |
node --test screens/bin/ | Run the /screens CLI test suite (screens.mjs's subcommand branches: plan, promote, guards, trust, affected, ...) |
node screens/bin/screens.mjs <plan|up|down|promote|marketing|index|trust|affected|migrate-layout|migrate-output|macos-export> | /screens' deterministic CLI: plan diffs the manifest against fingerprints, up/down handle isolation lifecycle, promote byte-hash-compares captured PNGs into the catalog (resolved via resolveScreensOutputRoot: config.output_dir when set, else ~/Developer/screens/<project>/, never inside the project) with an ImageMagick fuzz-tolerance fallback for rendering jitter, marketing renders store images, index regenerates the project's own index.html + catalog.json plus the top-level ~/Developer/screens/index.html, trust checks the confirmed command hash, affected --files <paths> maps changed files to manifest entry ids, migrate-layout moves an older output layout into the current device-class folders, migrate-output moves an existing project-local screenshots/ tree into the resolved central output root once, macos-export pulls macOS screenshots (delivered as XCTAttachments, since the macOS UI test runner is sandboxed and cannot write into the project directory) out of the .xcresult bundle via xcrun xcresulttool export attachments and renames them into .screens/.incoming/macos/ (platform-apple.md "macOS: attachment export") |
node screens/bin/ui-shots.mjs --url <url> [--out <dir>] [--login-url <url> | --form-login <url> <user-field> <pass-field> --creds-file <path>] | Self-check shots for web UI tasks: one browser launch, three PNGs (375 light, 375 dark, 1280 light), prints the final URL per shot and the title; exits 2 (SELF-CHECK: unreachable) when login fails or a shot lands on a login page. Reuses the pinned Playwright in ~/.cache/store-assets, never installs into the project. Run it through audit/bin/test-lock.sh --cmd. Tests: screens/bin/ui-shots.test.mjs |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.