All authors
Claude Skills by SkillfulAgents
github.com/SkillfulAgents18 skills0 installs0 views
- A11y AuditRun a runtime accessibility audit using axe-core via Playwright and review the resultsVotes: 0GitHub stars: 24
- Add Composio ProviderAdd a new Composio OAuth provider (toolkit) for connected accounts. Wires up the slug across providers list, proxy allowlist, scope map, scope descriptions, agent service discovery, system prompt, and downloads the icon.Votes: 0GitHub stars: 24
- Add Dashboard PolyfillAdd a new browser API polyfill that is auto-injected into dashboard iframes. Covers the polyfill implementation, injection wiring, testing (unit + E2E), and documentation for the dashboard builder agent.Votes: 0GitHub stars: 24
- Auth MigrationAdd a new Better Auth plugin or feature and generate the corresponding database migrationVotes: 0GitHub stars: 24
- Browser Transcript MiningMine a dump of production `web-browser` subagent transcripts (jsonl) for browser-harness friction — soft failures, workarounds, missing tools, misleading output. Deterministic stats + compact traces first, then Lane A (Opus reads whole sessions with screenshots, open rubric, discovery) and Lane B (Sonnet classifies compact traces against a closed taxonomy, prevalence). Use when a new transcript dump arrives, after shipping a harness fix to see if the metric moved, or when someone asks "where ...Votes: 0GitHub stars: 24
- Claude Prompt DriftDetect drift in the Claude system prompt that agent-container sends to Anthropic. Captures two axes — the bare `claude_code` preset (Anthropic baseline) and agent-container's fully custom wire request — at the SDK version pinned in `agent-container/package-lock.json`, then stores versioned snapshots in `<skill>/snapshots/` (gitignored). Because SuperAgent REPLACES the preset with its own system prompt, preset drift never reaches our sessions automatically — every pure-claude change must be tr...Votes: 0GitHub stars: 24
- Container BenchmarkRun the container startup benchmark to measure and validate the full agent-wake → dashboard-visible flow. Use before and after optimization work to catch regressions and compare timings.Votes: 0GitHub stars: 24
- Electron Cloud Interface ValidationStand up the three-node local stack (platform auth + proxy, a remote auth-mode deployment, this app as the Electron client) and run the gated live suite that proves the cloud-workspace chain — discovery → RFC 8693 grant → RFC 7523 exchange → persisted deployment token. Use when changing anything under cloud-workspace-*, the token-exchange endpoint, the platform grant/discovery endpoints, or before merging a PR that touches them.Votes: 0GitHub stars: 24
- Pr Demo VideoRecord a short demo video of a UI change with Playwright and embed it in the PR description or a PR comment with `gh --attach`. Use when a PR changes something a reviewer should see move, or when the user asks for a demo clip.Votes: 0GitHub stars: 24
- Pr ScreenshotsCapture light + dark screenshots of the UI a branch changes and attach them to the pull request with `gh pr create --attach` / `gh pr edit --attach`. Run before opening any PR that changes something a user can see (renderer components, styles, dashboards, wizard, settings). Also use to add screenshots to an existing PR.Votes: 0GitHub stars: 24
- Seed Test EnvScaffold a fresh isolated Superagent data dir under ~/Downloads, seed agents, copy curated settings from the prod install, and print the launch command. Use when manually testing a feature against a clean state without touching the real install.Votes: 0GitHub stars: 24
- Slack Chat ValidationRun the live Slack chat-integration validation suite — a real Slack account messaging a real bot, driving a real app instance and a real container, asserting on what actually lands in Slack. Use before and after any change to src/shared/lib/chat-integrations/, to the user-input request path that feeds it, or to the "can't do this in chat" notice.Votes: 0GitHub stars: 24
- Update Claude DepsReview and upgrade Claude Code CLI, Claude Agent SDK, and Anthropic SDK versions used in the projectVotes: 0GitHub stars: 24
- Update ModelsCheck and update to the latest Claude model versions across the codebaseVotes: 0GitHub stars: 24
- DashboardsCreate interactive web dashboards to visualize data and provide UI elements to the userVotes: 0GitHub stars: 24
- Site ApiFind a website's own JSON endpoints by recording its traffic in the browser, then write a script that calls them directly — so a task that repeats (a scheduled report, a widget, a price or availability check) makes one HTTP request per run instead of driving the browser. Use when asked to "find the API" of a site, "turn this site into a script", or when the same site will be read on a schedule.Votes: 0GitHub stars: 24
- WidgetsGive an artifact a small glanceable home-screen widget (one number, next event, a few KPIs) that refreshes itself from a scriptVotes: 0GitHub stars: 24
- Workflow AuthoringReference for writing a Workflow tool script (script API and gotchas, resume, quality patterns, worked examples). Load before authoring a script for a workflow the user already opted into; it does not itself authorize running one.Votes: 0GitHub stars: 24