agent-memory
claudecodexRecall and capture for a self-hosted Agent Memory server (needs the `amem` binary on PATH).
1 MCP server · hooks
Claude Code and Codex plugins their authors filed under "productivity". All categories.
Recall and capture for a self-hosted Agent Memory server (needs the `amem` binary on PATH).
1 MCP server · hooks
Recall and capture for a self-hosted Agent Memory server (needs the `amem` binary on PATH).
1 MCP server · hooks
Install, configure, and use the DFOS protocol CLI: identities, content chains, credentials, services, and relays.
1 skill
영어 PDF/원서를 자연스러운 한국어(존댓말·핵심용어 병기)로 병렬 3단계 번역하고, 에디토리얼 단행본 PDF로 디자인까지 한 번에. 'PDF 번역', '원서 번역', '번역해서 책으로' 요청에 사용.
1 skill
Minimum Sufficient Work: do all necessary work and nothing beyond it.
1 skill
Optional session-boundary reinforcement for the MSW kernel.
hooks
Bound agent work with Available Work Time, Closeout Grace Period, and a hard stop.
1 skill
A single source of truth for deferred work: ONE file per project (UNFORGET.md), four sections of 10-column rating tables, and a release-cycle Target column that promotes items toward shipping. Format v2 adds machine-readable status tokens (a done isn't done until it's verified), a ledger registry, an integrity-lint verify gate, a deferral gate that makes deferring cost something, atomic child ledgers, self-maintaining recall, and companion-skill handoffs. Backward compatible with v1 ledgers. Includes HTML ledger views (top 10 by urgency, quick wins, ship blockers, or your own mix) that say what they left out, never hide a ship blocker, and can read your project's own rating words.
1 skill
Skills for editorial craft on prose written for people to read: reports, proposals, updates, and documentation. Its first three skills are in development — slop_catch, which flags AI writing tells in a draft; ghost_writer, which carries per-genre rules for writing and editing strong prose; and language_humanizer, which delivers prose the intended reader understands on the first read.
1 skill
Translate a novel/epub end-to-end on the coding agent (Claude Code / Codex): parse, build a locked glossary, translate chapter by chapter (names preserved, title 3-way, term consistency), optional multi-agent parallel mode, write back and export. Self-contained Python pipeline (vendored parse/export for epub/txt/docx/xlsx/… — no AiNiee repo); pip install a few light libs. AINIEE_REPO optional, only for PDF/Office.
1 skill · 18 commands
Code review, fixes, commits, release flows, and linting for development workflows
7 skills · 2 agents · hooks
Research, docs lookup, utility routing, skill authoring, instruction review, and agent configuration audits
7 skills · 3 agents
Idiomatic development across C# /.NET, Go, Java/Kotlin, Python, Rust, TypeScript, shell, and web
8 skills · 1 agent
Strict, routed architectural guardrails for AI coding agents, powered by ParselFire Core.
2 skills
Name a paper or hand over a PDF: Trace finds it on arXiv or by DOI (bioRxiv, medRxiv, PubMed Central, ACL Anthology), grounds every claim in a page and quote, checks each quote against the text of its page, and builds a local interactive site that teaches it: primers, step-by-step derivations, playgrounds that ask for your prediction first, common misreadings and an evidence-linked quiz. Check your own explanation of a section against the evidence, see which concepts you already studied in your other papers, get a reading order for your library with the papers you saved to read later, your study progress, your work time against last week and your notes as Markdown or for Obsidian, rewrite one section with the evidence locked, publish a shareable link, and export a report, slides, a Jupyter notebook, a citation or Anki flashcards.
1 skill · 1 MCP server
Generate AI videos, images, voice, and use AI apps via the Zyka CLI. 40+ models in one place.
2 skills
Generate interactive HTML replays and GitHub PR artifacts from AI coding sessions
1 skill
Adds the unified okf-harness skill for workspace setup and daily maintenance through the pinned runtime launcher.
1 skill
AST-based code reducer for minimizing context size during debugging.
1 skill
Build a multi-market TikTok Shop video production line without learning traditional editing software.
1 skill
Behavior and workflow guardrails with knowledge accrual for Claude Code. 12 skills and 58 specialized agents covering problem-solving, work tracking, systems engineering, orchestration, adversarial quality review, secure engineering, offensive security, and transcript parsing.
32 skills · 2 commands · hooks
Autonomous Development Orchestrator for Claude Code. Spec → Plan → Tasks → 100% Working App. Multi-agent pipeline with zero manual intervention.
6 skills · 8 commands · 13 agents
Queue-based personal workflow assistant. Captures decisions, commitments, ideas, and session data via ultra-fast hooks (< 50ms), processes them in a background worker with AI extraction, and syncs to Obsidian vault. Replaces my-workflow with zero-latency capture architecture.
3 skills · 11 commands · 3 agents · hooks
Productivity & agentic office agent: 10 skills, 4 agents. Context loading, delegation, daily digest, executive briefs, meeting intelligence, progress tracking, task intelligence, and workplace search.
10 skills · 4 agents
Plan and execute tasks from PRDs with hierarchical subtasks, dependency tracking, and project memories. Features dashboard visualization, autonomous batch execution, and persistent memory for constraints, decisions, and conventions.
2 skills · 8 commands · 1 agent
Manage an Obsidian vault as your development knowledge base and automatic work journal. Auto-captures commits, tasks, and Claude Code components with structured frontmatter.
1 skill · 9 commands · hooks
Discipline mode (staged planning, receipts, verify-against-source, adversarial review) + plan/kickoff commands + a plan-reviewer agent + an always-on SessionStart discipline injector + a Stop-hook autonomous loop (Layer 1, a deterministic gate) that can arm itself from a committed, content-hash-trusted .claude/tale-mode.json during a kickoff phase + a fail-closed plan-mode auto-approve for provably read-only shell (no more permission-dialog walls during investigation).
6 skills · 2 commands · 1 agent · hooks
Draft and revise prose for clarity, usefulness, tone, and a natural human voice.
2 skills
python-hwpx 프로젝트가 직접 유지보수하는 first-party HWPX 에이전트 스킬. 편집은 순수 Python이며 최종 시각 검증은 필요 시 한컴 오라클을 사용합니다.
1 skill · 1 MCP server
Back up, analyze, and reuse ChatGPT | Claude | Codex | Cursor | DeepSeek | Qwen | Openclaw data + skills + attachments locally
1 skill