Agent Plugins Marketplace

Recently added

Plugins grouped by the day they were added to the directory, newest first.

Sep 14, 2026

  1. Office document generation plugin for PPTX, DOCX, and XLSX workflows.

    Codex3 skills
  2. Media generation and manipulation for images, videos/GIFs, music, speech, sourcing, SVG, and data visualization via Gemini, D3.js, Mermaid, and Draw.io, with visual-planning and media-prompt-craft gates.

    Codex10 skills5 MCP servers
  3. Infrastructure plugin for Kubernetes, Istio, Helm, Terraform, Traefik, and auth workflows.

    Codex6 skills
  4. Developer tooling for read-only thinking and Q&A, git workflows, code hygiene, dependency management, and project documentation. Every skill is user-invoked only.

    Codex9 skills1 MCP server
  5. Design direction and creative guidance for styleguides, aesthetic strategy, typography, color systems, design review, and design systems. Requires media-plugin for asset generation.

    Codex4 skills
  6. Build narrated, edited project demos with deterministic app state; Playwright or OBS capture; ElevenLabs voiceover; measured timelines; Remotion rendering; and Final Cut Pro export with first-class motionVFX inventory, previews, titles, transitions, and effects.

    Codex8 skills2 MCP servers
  7. AI company planning plugin for Paperclip-style company, agent, and infrastructure design workflows.

    Codex3 skills
  8. Full research -> draft -> edit pipeline to a publish-ready piece; use for "write me an article on X end to end", "take this from idea to finished post". Use this when the user wants the whole pipeline, not a single stage.

    Claude Code1 skill
  9. writer0stars

    Drafts a piece from a brief (and optional dossier) — drafting ONLY, no research or editing pass; use for "draft this", "write a first draft from these notes", "turn this brief into a post".

    Claude Code1 skill
  10. wrap0stars

    Full session close-out ritual — commit, export conversation, teaching narrative, handoff, memory harvest; use for "wrap up", "close out", "export this conversation". For a plain .context/ checkpoint, prefer session.

    Claude Code1 skill
  11. Installs and operates the outcome-driven WBS toolkit as a folder's agentic planning and verified execution function; use for "set up WBS here", "spec this project", "run the WBS", or "update the WBS toolkit".

    Claude Code1 skill
  12. Derives a reusable voice profile from authentic writing samples so AI output sounds like the user; use for "doesn't sound like me", "write in my voice", "set up my voice profile".

    Claude Code1 skill
  13. Builds a hands-free voice-to-database capture chain (Siri to Shortcut to GPT to Sheets); use for "I never update my CRM", "talk to my database", "voice-to-sheet workflow".

    Claude Code1 skill
  14. Runs a folder's declared .aai/triggers.md standing intentions — compares each to last-run stamps in .aai/memory/, runs what's due, and can offer promotion to native hooks/cron on a capable harness; use for "check triggers", "run due triggers", "session-start due-check".

    Claude Code1 skill
  15. Trello-based agentic product pipeline turning an idea into a PRD via staged AI agents on a live board; use for "run the Trello pipeline", "idea to PRD on my board", "set up the TrelloAgents board".

    Claude Code1 skill
  16. Turns a raw transcript or session notes into a full content package (article, LinkedIn variants, summary, audio script); use for "turn this into content", "write something from this transcript".

    Claude Code1 skill
  17. to-issues0stars

    Breaks a plan, spec, or PRD into independently-grabbable tracer-bullet issues on the project's issue tracker; use for "turn this plan into issues", "break this PRD into tickets".

    Claude Code1 skill
  18. Writes one publishable teaching-block article from a working session — real prompts, real outputs, wrong turns, reasoning, diagrams, and rendered HTML; use for "write this up as a teaching block", "field guide this", "debrief this session", "turn this session into an article". Any domain. For the full look-over-my-shoulder bundle (README, brief, chat export, vault artifacts) use extract-codify-patterns instead.

    Claude Code1 skill
  19. Installs and operates SoloFactory, a local-first software factory (interview → PRD → subscription-backed build → gates → local deploy) as a folder's agentic function with each project a git repo under projects/ or the root; use for "set up solofactory here", "make this folder a software factory", "start the factory", "update my solofactory".

    Claude Code1 skill
  20. Installs and operates the software development factory, a governed SDLC controller that drives your own Claude Code or Codex CLI through plan, red, green and review against an approved spec, as a folder's agentic function with one git repository per product under projects/; use for "set up the software dev factory here", "make this folder a software factory", "approve a spec", "run the factory", "update my software-dev-factory".

    Claude Code1 skill
  21. Runs one skill's existing evals/evals.json — triggering cases against a fresh subagent, quality cases judged against prose pass criteria — and reports a PASS/FAIL table, no scores; use for "run the evals for X", "test this skill's evals", "does X still trigger right".

    Claude Code1 skill
  22. Audits a skill's SKILL.md frontmatter, catalog.yaml projection, references, and evals shape against the library schema, reporting ordered actionable findings; use for "audit this skill", "does X follow the schema", "lint X against the catalog", "check X before I publish it".

    Claude Code1 skill
  23. session0stars

    Runs session check-in/check-out against a project's .context/ store; use for "what should I work on", "resume work", "checkpoint", "session check-in/check-out". Requires (or offers to bootstrap) a .context/ directory.

    Claude Code1 skill
  24. Gathers sources into a structured research dossier; use for "research this topic", "find sources on X", "compile background". NOT for writing or editing the piece.

    Claude Code1 skill
  25. Scans one or more directories (no recursion) and produces a project inventory — markdown table plus a searchable, filterable HTML page — with an inferred one-line description and status per folder; use for "what have I been working on", "index my projects folder", "status of everything in this directory", "catalog these folders".

    Claude Code1 skill
  26. Creates a one-page project overview; use for "write a brief", "summarize this project", "project overview", "what is this project".

    Claude Code1 skill
  27. Gently coaches when the user writes in a non-English language mid-conversation — corrects, teaches, logs, returns to task; applies whenever the user code-switches as practice.

    Claude Code1 skill
  28. Generates non-obvious insight by colliding 400+ disciplinary paradigms via parallel sub-agents and an elimination tournament; use for "fresh angle on X", "non-obvious take", "multiple lenses".

    Claude Code1 skill
  29. Orchestrates a live human-steered Claude/Gemini/Codex debate against one shared markdown file; use for "three-pane terminal", "run the models against each other" — the live workflow, not the prompt protocol.

    Claude Code1 skill
  30. Turns meeting transcripts into newsletter-style topic recaps with research add-ons and optional Notion publishing; use for "recap this meeting", "session report", or any pasted VTT/call transcript.

    Claude Code1 skill
  31. Selects and applies storytelling frameworks for Leaderize's mid-level-leader audience across content stages and live sessions; use for Leaderize content or session design.

    Claude Code1 skill
  32. Multi-phase latent-space traversal that maps the probability landscape before answering; use for "/lsc", "run the cartographer", "don't give me the median answer".

    Claude Code1 skill
  33. Guides design and deployment of a personal AI knowledge base from accumulated expertise assets; use for "talk to my own knowledge", "build a private ChatGPT on my material".

    Claude Code1 skill
  34. Guided live discovery of a professional's irreplaceable competitive edge against AI-first competition; use for "find my edge", "what's my moat", "how do I compete with AI".

    Claude Code1 skill
  35. ireport0stars

    Generates styled, interactive HTML research reports from a topic or notes; use for "interactive report", "HTML report on X", "deep research document".

    Claude Code1 skill
  36. Hardens a contrarian claim by forcing it to survive repeated self-attack until it is simultaneously novel AND true; use for "give me a hot take on X", "pressure test this take", "is this angle actually original".

    Claude Code1 skill
  37. Interviews the user to design a stateful-agent-harness ProcessSpec (a state machine of agents, skills, gates, artifacts); use for "build an agent process for X", "spec out an agent pipeline". NOT for code work-breakdowns (wbs-prd).

    Claude Code1 skill
  38. grill0stars

    Converges on a decision or spec via one high-leverage question at a time with a decisions ledger, then executes; use for "grill me", "interview me", "ask me questions one at a time".

    Claude Code1 skill
  39. graphify0stars

    Turns any folder of files into a persistent, queryable knowledge graph with community detection; use for codebase/architecture questions, "graphify this", or any content query when graphify-out/ exists.

    Claude Code1 skill
  40. Two-phase Gmail inbox sweep via the gmail MCP — batch-archive junk, then triage real emails one at a time; use for "check my email", "clean up my inbox".

    Claude Code1 skill
  41. General Gmail management across accounts via the gmail MCP (search, read, send, label, archive, drafts); use for "send an email", "find that thread", or mailbox actions beyond triage.

    Claude Code1 skill
  42. Builds a GEO/AI-citation authority strategy and Canon/Frameworks/Diagnostics content architecture; use for "GEO strategy", "get cited by AI", "GEARS".

    Claude Code1 skill
  43. Sends any email to GEARS licensees (gears_owners) via Resend, drafting subject/body with the user if not already given, straight from raw HTML/text with no dashboard-editor round trip; use for "send a gears broadcast", "email the licensees", "notify gears owners", "send this to gears_owners".

    Claude Code1 skill
  44. gauntlet0stars

    Runs ambitious work through explicit criteria, independent critique, and evidence-based revision loops.

    Claude Code1 skill
  45. Validates docs/onboarding/workflows by spawning a zero-context agent to walk through them cold, then fix-and-respawn; use for "fresh eyes", "does the onboarding work", "test the docs".

    Claude Code1 skill
  46. Packages a working session into a shareable look-over-my-shoulder artifact bundle; use for "package this session", "codify this session", "turn this conversation into a teaching resource".

    Claude Code1 skill
  47. Rewrites and augments a prompt to surface implicit constraints and blind spots, then answers the enriched version; use for "make this prompt better", "I keep getting shallow answers".

    Claude Code1 skill
  48. Nine-step framework turning tacit expertise into branded, publishable IP (UP to IP); use for "turn how I think into a framework", "package my methodology", "make my process teachable".

    Claude Code1 skill
  49. Senior-editor orchestrator running specialist contributors (strategist, angle, drafter, adversary, headline) over a shared blackboard; use for "run the editorial desk", a critiqued-and-revised opinion/POV essay, or a sharp argument piece with editorial review.

    Claude Code1 skill
  50. editor0stars

    Surgically edits an EXISTING draft; use for "edit this draft", "polish this", "tighten this article". Requires a draft to already exist.

    Claude Code1 skill