Front-end-first delivery for web, React Native and native mobile: build the app on a typed mock API layer, run tester feedback rounds until the design settles, derive the backend contract and a PRD from the settled mocks, then swap mocks for live calls in probed waves with a live mock registry; plus a QA loop from test plan to tester feedback, validation, questions and fix plan.
Plugin mock-first-frontend-release from marketplace prabhg-plugins in prabhg/skills-marketplace. Commands read from the repo's manifest — 14 days ago.
fe-to-backend-specDerive the backend contract from a front end that already works on a mocked API layer: inventory every operation from the mock layer and registry, then write the backend contract (GraphQL SDL or OpenAPI, events/realtime topics, permissions per operation, data model with explicit id kinds, derived fi
feedback-roundsRun the tester-and-designer feedback loop on a front end that is built on mocks (or partly live): ship preview builds or over-the-air updates to testers, collect feedback (tester notes, annotated screenshots/PDFs, designer mock files), triage it into a numbered round work list, implement and re-ship
mock-first-appBuild a fully working front end (web, React Native/Expo, or native iOS/Android) on a typed, per-operation mock API layer before any backend exists, so testers can use it and each mock later swaps for a live call without touching screens. Sets up: per-operation mock/live bodies, one-intent mutations
mock-to-live-integrationReplace a front end''s mock API operations with live backend calls, safely and in waves: refresh the schema snapshot, inventory every operation against the live schema, find never-called backends, probe each new operation before designing against it, plan waves with disjoint write scopes, route back
qa-loopRun the QA loop on an app testers use from preview builds, where feedback comes back against a test plan (case ids, Pass/Fail marks, notes): write the per-screen test plan testers follow (a private page they print or save as PDF), extract their filled-in feedback (handwritten or typed PDF annotation
Compact a long session without losing it: save goal, verbatim decisions, state and next steps (durable facts to memory, working state to a token-named file), then a SessionStart(compact) hook re-injects the file after /compact and deletes it.
1 skill · hooks
Action-first output shaping: next action first, numbered steps, no tangents, visible progress.
1 skill · hooks
Carry a project across Claude account logins on one machine: before switching, capture work in flight, live sessions, git state and everything tied to the account into a local handoff; after switching, verify, re-prime and resume. Any number of accounts, repeated round trips.
1 skill
Build a single-file HTML architecture review doc for a team: layered system map, one pre-rendered sequence diagram per user action, services and ownership, APIs by domain, tiers, access rules, environments, security, rollout and decisions.
1 skill
Run long unattended multi-agent builds from one session: lean-context orchestrator, tickets with DoD, Haiku/Sonnet/Opus 5.5+ subagents by risk, review before every merge, wave merges to dev, green builds, cleanup, one resumable state file, owner-set handoff threshold (default 60%) to versioned successors, coordinated mode under orchestration-coordinator.
1 skill
One owner-facing coordinator over several autonomous-orchestrator sessions: disjoint repo lanes, owner rulings recorded verbatim, batched owner questions, cross-lane bug and contract routing, every owner command numbered and printed once, context-threshold handovers, minimal text in every session.
1 skill
Talk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
tasteskill 2.5.1: professional company websites at the figma.com and x.ai level, with one section per screen, real photographs, real product UI, effects kept for tech and gaming pages, and a real-browser checker and finish gate.
1 skill
Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications
5 skills · hooks
Cross-session memory and token savings for coding agents.
6 skills · 1 agent · 1 MCP server · hooks
Package any website or local web build into a lightweight desktop app with Pake CLI.
1 skill
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.