Agent Plugins Marketplace

Recently added

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

Sep 6, 2026

  1. tandem1stars

    Bidirectional bridge between Claude Code and Codex (ChatGPT): code review, adversarial review, task delegation, background jobs, models consulting each other, and image generation via Codex's built-in tool. Runs on your existing ChatGPT subscription, no API keys.

    Claude Code2 MCP servers
  2. whetstone33stars

    Engineering-discipline skills for planning, debugging, review, verification, and implementation in Codex.

    CodexClaude Code32 skills1 MCP server
  3. unknowns1stars

    Know your unknowns — an operational loop for finding and managing the gaps between the map (your plan) and the territory (the actual code and domain). Provides blind-spot investigation (blindspot), domain primers (teach-me), interviews, divergent prototypes, solution brainstorming, reference analysis, tweak-likelihood-ordered planning (plan), deviation notes (notes), independent verification, buy-in documents, and comprehension quizzes as skills/agents/hooks. Based on Thariq Shihipar's (Anthropic) 'Field Guide to Fable' talk and its 'Know your unknowns' example collection.

    Claude Code11 skills
  4. spill-guard0stars

    Blocks secrets and PII before they reach the API: scans the prompt, a Read path, a Bash command, and the files each one would open, then denies the call rather than letting it cross.

    Claude Code
  5. trustme0stars

    e2e fixture: hooks + command tool surfaces

    Agent Plugins
  6. groq-and-review0stars

    Groq provider + a review skill + a lint hook

    Agent Plugins1 skill1 MCP server
  7. warned-plugin0stars

    warned-plugin is an open-source Agent Plugins plugin indexed from GitHub. Review its manifest and source before installing.

    Agent Plugins
  8. ext-ignored-plugin0stars

    ext-ignored-plugin is an open-source Agent Plugins plugin indexed from GitHub. Review its manifest and source before installing.

    Agent Plugins
  9. bare-plugin0stars

    plugin.json only, no skills/, mcp.json or dev.instagent/

    Agent Plugins
  10. liveplug0stars

    live e2e fixture: command tool + hooks + skill + slash command

    Agent Plugins1 skill
  11. bundled0stars

    Instagent bundled plugin: built-in provider definitions

    Agent Plugins
  12. session-orchestrator49stars

    Loop engineering for AI coding agents — turn ad-hoc sessions into a repeatable research → plan → wave-execute → close loop with verification gates. Runs on Claude Code, Codex CLI, Cursor, and Pi.

    CodexClaude CodeAgent Plugins43 skills1 MCP server
  13. superpowers-orchestrator3stars

    Structured workflow skills for Codex — systematic debugging, TDD, brainstorming, code review, and 30+ expert workflows. Skills work on all platforms. Lifecycle hooks (routing, safety, reminders) require separate setup on macOS/Linux.

    CodexClaude Code28 skills
  14. livespec0stars

    Codex Driver for livespec: thin skill bindings exposing the eight livespec operations (seed, propose-change, critique, revise, doctor, prune-history, next, help) as Codex plugin skills. The harness-neutral prose and the reference spec-side CLIs ship with livespec core (install livespec@livespec alongside this Driver).

    Codex8 skills
  15. notrest-free0stars

    PART OF ATLAS → the notrest harness ships to Atlas customers from v4.8.0. This entry exists only to tell existing installs where the harness went and how to ask for access. Versions through v4.7.1 were released under the MIT License and remain so.

    Claude Code
  16. zerocool0stars

    Gibson for Claude Code. The model stays on your Claude login.

    Claude Code1 MCP server
  17. deep-research0stars

    Grill-to-brief scoping, parallel DuckDuckGo web and arXiv paper divers, citation-gated Markdown reports.

    CodexClaude Code2 MCP servers
  18. weavatrix4stars

    Evidence infrastructure for AI software agents: 47 read-only repository intelligence tools through MCP.

    CodexClaude Code1 skill1 MCP server
  19. software-dev0stars

    Glue over superpowers and mattpocock/skills for Claude Code and Codex: narrowed brainstorming, the repository scaffolder, the consistency audit, diagnosing-bugs, and finding-duplicate-functions.

    CodexClaude Code5 skills
  20. sensemaking0stars

    Skills shared by software-dev and research-vault: rethink-audit and adhd.

    CodexClaude Code2 skills
  21. junto0stars

    Workflow engine for Claude Code with durable state and evidence-based quality gates. No binaries, no runtime downloads, and writes only under the project's .junto directory.

    Claude Code1 MCP server
  22. skillcrossroads1stars

    Audit Claude Code artifacts from inside Claude Code: the audit-skill runs the free skillcrossroads CLI to grade skills, subagents, slash commands, MCP configs, and plugins with an evidence-cited file:line scorecard, then walks the ranked fix list until the grade stops improving.

    Claude Code1 skill
  23. deploy-tools1stars

    Deployment automation for monorepos: preview-URL commands, env-var sync, and a pre-deploy checklist hook.

    Claude Code
  24. skillcrossroads1stars

    Audit your Claude Code skills from inside Claude Code: installs the audit-skill, which runs the skillcrossroads CLI (the grader for skills, subagents, slash commands, MCP configs, and plugins) to produce an evidence-cited, file-and-line scorecard and walk the ranked fix list until the grade stops improving.

    Claude Code
  25. architecture-doc-plugin0stars

    A production-grade codebase architecture documentation plugin for Claude Code. Features AST-level dependency graph extraction, TypeScript path alias resolution, deterministic API endpoint discovery, Mermaid.js diagram generation with syntax validation, interactive HTML previewing, and automated CI/CD documentation workflows.

    Claude Code1 skill
  26. iec62304-writer1stars

    IEC 62304 class A technical-file pipeline — SRS, SDS, STD, traceability, safety (ISO 14971 §C.2 causal chain + §7.2 control hierarchy + §7.4 quantitative residual + §7.5 cascade), production, cyber (IEC 81001-5-1) and usability (IEC 62366-1 — UEF + Summative Evaluation + IECEE checklist, platform-rich and clinical-narrow modes) risk analyses, incremental update, RAQA-ready exports (SRS + SDD + STP + STDR + STR + Risk Analysis Report + Risk Table xlsx + UEF + USE + UEF-Annex1), and a submission contract (normative vs internal sections, requirement kinds, parameter registry, OTS registry, cybersecurity architecture views) enforced by a gated --release export. Reproduces locally the useful features of Matrix Requirements.

    Claude Code22 skills
  27. undervalued-now0stars

    One paragraph, no capabilities, and its own web and shell. A benchmark of the model rather than of a methodology — the prompt is the question, and the forward return is the answer.

    Claude Code
  28. prudent-allocator0stars

    Top-down and risk-budget-first. Starts from the book rather than from the asset, sizes every judgement by what a bad outcome costs the portfolio, and acts only when the exposure is wrong.

    Claude Code
  29. evidence-gated0stars

    One evidence-gated methodology run as three market flows: a Korean sleeve, a US sleeve, and the allocator that sets their budgets and proposes the cross-market rebalance. The flows are subagents of one manager; only the orchestrator submits.

    Claude Code14 skills1 MCP server
  30. earnings-drift-watcher0stars

    Judges one thing only: whether the market has finished reacting to a reported surprise. Compares the reported figure against what was expected, then against how far the price has already travelled, and mostly concludes that it is too late.

    Claude Code
  31. basic-investor0stars

    FinRobot's Data-CoT → Concept-CoT → Thesis-CoT, extended with Portfolio Context and a Decision stage. The reference ManagerPackage.

    Claude Code
  32. atlas-trend-us0stars

    Time-series momentum across a fixed basket of US-listed global ETFs. Holds only what is trending in its own price, sizes what it holds by realised volatility, and goes to T-bills when nothing is. One monthly rebalance, or nothing.

    Claude Code2 skills
  33. atlas-trend-kr0stars

    Time-series momentum across a fixed basket of Korea-listed ETFs, in won. Holds only what is trending in its own price, sizes what it holds by realised volatility, and parks in CD-rate funds when nothing is. One monthly rebalance, or nothing.

    Claude Code2 skills
  34. atlas-trend-crypto0stars

    Time-series momentum on crypto spot, in dollars. One or two assets and cash, because a basket of coins is one position wearing ten names. Holds only while the trend is up, sizes the exposure by realised volatility, and goes to cash when it is not. One monthly review, or nothing.

    Claude Code2 skills
  35. ai-hedge-fund-value0stars

    A port of ai-hedge-fund's deep-value pod: Graham, Buffett and Munger read the same point-in-time fundamentals independently, and one reconciliation weighs their convictions. Graham's view counts double, as the source's own strategy file specifies.

    Claude Code
  36. sec-overlay0stars

    Agentic security-audit harness: runs SAST, investigates candidates with multi-agent gates, and emits SARIF + Markdown reports.

    Claude Code1 skill
  37. design-dna1stars

    Cross-host website production with specific, source-authorized visual decisions and no producer-authored design, visitor-serving copy, rendered verification, and opt-in Connected Public Experience.

    CodexClaude Code1 skill
  38. makoto45stars

    Holds Claude Code to its word — every claim checked against the agent's own logged record; faked checks blocked, not warned. Two-tier catalog (36 pre-checks, 26 Stop gates), each shipped at measured zero false positives.

    Claude Code
  39. product1stars

    Keeps a repository's beliefs, risks and outcomes in contact with reality. Challenges and proposes; never decides meaning.

    Claude Code2 skills
  40. clickai-codex1stars

    23 Click AI skills for Codex desktop, adapted for Astra workflows: planning, research, writing, implementation, independent review, and checkpoint recovery.

    Codex
  41. clickai-skills1stars

    Portable skills for AI agents — a multi-model advisory board, a sixteen-skill team-workflow pack for tracked multi-session development, a media-ingest skill that turns recordings and URLs into evidence, a pair of writing references: documents agents consume, and pages humans read — and a plain-language decoder for agent messages that didn't land.

    Codex
  42. ca145stars

    Orchestration layer for Claude Code. Routes every intent through gated skills and reviewer agents, drives spec-driven TDD, mechanically enforces the commit and audit-trail gates, decides via SMARTS, and keeps an append-only audit trail. Requires Python 3 on PATH. Dormant until you opt a repo in; run /ca:init to activate.

    Claude Code23 skills
  43. ca-sandbox145stars

    Locally-hosted Codespace equivalent for codeArbiter. Pulls an untrusted repo into an ephemeral, isolated Docker container with no host-filesystem access and configurable egress, caches dependencies by content hash, then tears the box down. Requires Docker and nixpacks on PATH.

    Claude Code1 skill
  44. ca-codex145stars

    Governance kernel for OpenAI Codex CLI: the generated codeArbiter governance surface plus enforcement hooks (persona injection, blocking pre-exec and pre-write gates, append-only audit trail), sharing one .codearbiter/ store with the Claude Code sibling plugin. Standalone: opt a repo in with ca-init; enforcement stays dormant until .codearbiter/CONTEXT.md carries 'arbiter: enabled'. Requires Python 3 and Codex >= 0.143.0. CI continuously verifies, through a real Codex host at 0.143.0 and 0.145.0, that the plugin installs, reads back enabled, and ships every hook script it declares; an advisory lane tracks npm latest for upstream drift. Hook FIRING - live persona injection and live blocks inside a turn - is verified by hand per release against docs/codex-parity-testing.md, because a turn needs a model and a provider credential cannot gate fork pull requests.

    Codex36 skills
  45. codex-skill-resource-probe145stars

    Offline fixture for Codex skill-resource resolution characterization

    Codex1 skill
  46. linkedin-ads-mcp-kiin0stars

    LinkedIn Ads MCP — powered by Kiin. Read performance and write campaigns, budgets, targeting, creatives, and audiences (every write is preview-then-confirm).

    Agent Plugins1 skill1 MCP server
  47. workspace0stars

    Write what a workplace runs on: Korean internal communication, implementable technical specifications, and MCP access to Notion.

    Agent Plugins3 skills1 MCP server
  48. saju0stars

    Read a birth chart the way a myeongri researcher would: calculate the four pillars and luck cycles from solar terms with the working shown, then analyse the chart school by school — Ziping Zhenquan, Ditian Sui, Qiongtong Baojian, strength-balance, modern ten-gods psychology — and compare where they agree and where they do not.

    Agent Plugins1 skill
  49. research0stars

    Gather source material an agent cannot reach on its own: web search and AWS documentation, with skills for Agent Studio builtin document creation, inspection, and editing.

    Agent Plugins2 skills1 MCP server
  50. engineering0stars

    Get a change reviewed and out the door: a review that reads the diff instead of the repository and runs only the lenses its signals call for, PR descriptions built around intent rather than a file list, and the plain voice both of them are written in.

    Agent Plugins3 skills