OthmanAdi
UserPersistent file-based planning for AI coding agents and long-running tasks. Crash-proof markdown plans, session recovery after /clear and compaction, per-turn re-injection against context rot, deterministic completion gate. Manus-style. Install from npm, the Claude Code plugin marketplace, or npx skills. Codex, Cursor, OpenCode, 60+ agents.
Categories
Indexed Skills (25)
pi-planning-with-files
Persistent file-based planning for multi-step AI-agent work. Keeps task_plan.md, findings.md, and progress.md on disk; lifecycle hooks inject selected project planning context. Automatic recovery reads project planning files only. Explicit session-catchup.py --metadata reads same-project local agent session records and emits aggregate counts only; --replay may emit bounded nonce-framed excerpts. Optional gated mode can request continuation only when the host supports it and never runs commands declared in Markdown. The skill has no network upload path. Use for research or work needing 5+ tool calls.
planning-with-files
Persistent file-based planning for multi-step AI-agent work. Keeps task_plan.md, findings.md, and progress.md on disk; lifecycle hooks inject selected project planning context. Automatic recovery reads project planning files only. Explicit session-catchup.py --metadata reads same-project local agent session records and emits aggregate counts only; --replay may emit bounded nonce-framed excerpts. Optional gated mode can request continuation only when the host supports it and never runs commands declared in Markdown. The skill has no network upload path. Use for research or work needing 5+ tool calls.
planning-with-teams
Manus-style context engineering for Agent Teams. Coordinate multiple Claude Code instances with shared planning files. Use when complex tasks need parallel work (code review, debugging, feature development). Requires CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1.
chronos
Gives coding agents temporal awareness by reading the session transcript the host already writes, so the agent stops guessing about time and stops repeating commands that just failed. Activates whenever an agent reasons about recency, elapsed time, retry windows, deploy cooldowns, memory or file staleness, idle detection, or any "when / how long ago / how long did that take" question. Required for autonomous, parallel, or long-running agents.
plandeck
A live visual Kanban board and continuity layer for long-running agentic tasks. Nobody wants to read a markdown plan or stare at raw HTML while an agent grinds for an hour, so Plandeck turns the plan into a board that organizes itself. Use it when a task is multi-step, long-running, or needs visual planning. Break the work into cards, declare which card depends on which, and Plandeck computes the critical path, auto-promotes cards to Ready as their dependencies clear, rolls up story-point estimates into an honest percent complete, and names the one next action. Plans, observed transitions, and last-known-good snapshots stay as plain files on disk, so they survive /clear and context resets. After a reset, run plandeck next to recover the single next move and its recent history. If the plan stops parsing, run plandeck doctor before using git. Renders a live board in the browser over SSE. Triggers on plan a project, break down a task, kanban board, task board, visual plan, long-running agent, agentic task, track
openui-forge-anthropic
OpenUI generative UI with Anthropic Claude SDK backend. Stream conversion to OpenAI NDJSON format.
openui-forge-csharp
OpenUI generative UI with C# ASP.NET Core Minimal API backend. Direct OpenAI API SSE streaming via HttpClient on .NET 10.
openui-forge-elixir
OpenUI generative UI with Elixir Phoenix backend. Chunked SSE streaming via Plug.Conn and Req.
openui-forge-go
OpenUI generative UI with Go (net/http) backend. Direct OpenAI API streaming via HTTP.
openui-forge-java
OpenUI generative UI with a Java Spring Boot (WebFlux) backend. Streams the OpenAI API directly via WebClient as SSE.
openui-forge-langchain
OpenUI generative UI with LangChain/LangGraph backend. Supports ChatOpenAI and ChatAnthropic.
openui-forge-openai
OpenUI generative UI with OpenAI SDK backend. Streaming chat completions with gpt-5.5 (or any current OpenAI-compatible model).
openui-forge-php
OpenUI generative UI with a PHP (Laravel 13.x) backend. Forwards OpenAI's SSE stream verbatim via response()->stream().
openui-forge-python
OpenUI generative UI with Python FastAPI backend. OpenAI and Anthropic SDK variants.
openui-forge-ruby
OpenUI generative UI with a Ruby on Rails backend. SSE streaming via ActionController::Live, forwarding the OpenAI API stream with Net::HTTP.
openui-forge-rust
OpenUI generative UI with Rust Axum backend. Async SSE streaming with reqwest and async-stream.
openui-forge-vercel
OpenUI generative UI with Vercel AI SDK. streamText, toUIMessageStreamResponse, and tools support.
openui-forge
Build generative UI with OpenUI — any LLM provider, any backend language. Scaffold, integrate, validate.
pr-perfect
Create a GitHub PR with an engineered description that matches the team's established format — numbered problem-fix sections, technical tables, file-by-file breakdowns, commit lists, and verification checklists. Use this skill whenever the user asks to create a PR, ship code, prepare a merge request, write a PR description, or says anything like "ship to boss", "pr perfect", "make a PR", "prepare PR for merge", "create pull request", "engineer the PR", "PR description", or "ready to push". Also trigger when the user has finished a batch of work and says things like "let's wrap this up", "time to merge", "push this to main", or "get this ready for review". The skill handles everything: commit analysis, description generation, branch creation, and PR submission via gh CLI.
gmira-a11y
Use when auditing a built page for contrast, keyboard operation, semantics, and canvas fallback. Also use when text sits on a colored or image background and nobody measured it, when the focus outline was removed, when tab order jumps around the screen, when a div is being used as a button, when a canvas or WebGL layer carries text or controls, when a form field has no label, or when reduced motion only slows an animation down. Runs at audit time and not during design, because an accessibility reminder while designing produces timid underdesigned output. Ships a runnable DOM audit that computes contrast from computed styles rather than eyeballing it.
gmira-arsenal
Use when installing, repairing, or taming any shadcn registry component (componentry, canvas-ui, bklit, ncdai, kibo-ui, react-bits, soundcn). Also use when an installed component fails to build, imports a package that is not there, renders the wrong colors, shows a blown-out glow, ignores its own props, or looks like the library's demo instead of the brief. Covers registry search, install, import repair, dependency repair, welded-constant auditing, and palette verification.
gmira-ship
Use when a web surface is finished and needs to be built, staged, or deployed, and when preparing a release of the work. Covers the production build, bundle and asset checks, metadata and social cards, robots and sitemap, redirects, environment separation, and the staging handoff. Also use when someone asks to publish, push live, or deploy, because this skill holds the rule that deployment needs an explicit human go.
gmira-slop
Use when a built page needs the audit for the visual tells no detector catches, or when someone says a page looks generic, templated, bland, stock, or "AI-made" and cannot say why. Run it before showing any surface to the user, and again after a redesign pass. Detects uniform section rhythm, the three-column reflex, one crop ratio everywhere, accent-as-confetti, identical section entrances, the icon-plus-heading-plus-two-lines feature card, the ghost card, grids that truncate real content to stay even, decorative chrome standing in for data, a full-bleed effect turned down until it does nothing, and single-registry house style. Produces a findings table with severity that says what should happen, plus a specificity score per section.
gmira-verify
Use before handing any web surface back to a user, and whenever a page needs checking at real viewport sizes. Runs Playwright across five viewports, captures screenshots, collects console errors, measures computed contrast and text overflow, exercises pointer-driven effects with a real mouse, and checks reduced-motion and canvas-deleted fallbacks. Also use when something looks right in one browser size and wrong in another, when an effect "works" but nothing visible happens, or when a build passed lint but nobody has actually looked at it.
gmira
Use when building, redesigning, or reviewing a website or web surface that must not look generated. Covers landing pages, e-commerce (PLP, PDP, cart, checkout), car and vehicle inventory sites, course and cohort school pages, portfolios, dashboards, and marketing sites. Handles visual direction, WebGL and canvas effects, shadcn registry components, scroll choreography, typography, color, motion, empty and error states, accessibility, performance budgets, and multi-viewport visual verification. Also use when a page came out bland, templated, or "AI-looking" and needs a real direction, or when an installed registry component renders wrong.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.