What AI agent plugins do: a topic map
Catalog read · 1,183 plugins · 15 topics
746 of the 1,183 packaged plugins in the HOL catalog carry tags, and 539 of those fall into at least one of 15 topics. The largest topic is planning and workflow with 89 plugins, then memory and context with 88.
The topic map
Each tile is a topic, sized and shaded by how many plugins it holds. A plugin with tags in two topics counts in both. Pick a tile to jump to that topic’s highest-scored plugins.
- Planning and workflow89 plugins
- Memory and context88 plugins
- Code review77 plugins
- Docs and writing76 plugins
- Security and audit66 plugins
- Multi-agent orchestration57 plugins
- Testing and TDD57 plugins
- Images, audio and video55 plugins
- Design and frontend46 plugins
- SEO and marketing46 plugins
- DevOps and Git45 plugins
- Research and search40 plugins
- Debugging and observability39 plugins
- Payments and finance31 plugins
- Cost and token usage24 plugins
Which agents each topic supports
The share of each topic’s plugins that declare Claude Code, Codex and Cursor. A plugin can declare all three, so the shares in a row can add up to more than 100%.
Claude CodeCodexCursor
- Planning and workflow
- Claude Code
- 69% · 61
- Codex
- 75% · 67
- Cursor
- 25% · 22
- Memory and context
- Claude Code
- 78% · 69
- Codex
- 66% · 58
- Cursor
- 26% · 23
- Code review
- Claude Code
- 92% · 71
- Codex
- 70% · 54
- Cursor
- 29% · 22
- Docs and writing
- Claude Code
- 79% · 60
- Codex
- 63% · 48
- Cursor
- 20% · 15
- Security and audit
- Claude Code
- 76% · 50
- Codex
- 68% · 45
- Cursor
- 24% · 16
- Multi-agent orchestration
- Claude Code
- 79% · 45
- Codex
- 65% · 37
- Cursor
- 30% · 17
- Testing and TDD
- Claude Code
- 81% · 46
- Codex
- 72% · 41
- Cursor
- 28% · 16
- Images, audio and video
- Claude Code
- 66% · 36
- Codex
- 82% · 45
- Cursor
- 26% · 14
- Design and frontend
- Claude Code
- 87% · 40
- Codex
- 67% · 31
- Cursor
- 39% · 18
- SEO and marketing
- Claude Code
- 87% · 40
- Codex
- 85% · 39
- Cursor
- 24% · 11
- DevOps and Git
- Claude Code
- 73% · 33
- Codex
- 71% · 32
- Cursor
- 24% · 11
- Research and search
- Claude Code
- 78% · 31
- Codex
- 65% · 26
- Cursor
- 25% · 10
- Debugging and observability
- Claude Code
- 85% · 33
- Codex
- 72% · 28
- Cursor
- 23% · 9
- Payments and finance
- Claude Code
- 71% · 22
- Codex
- 74% · 23
- Cursor
- 36% · 11
- Cost and token usage
- Claude Code
- 88% · 21
- Codex
- 63% · 15
- Cursor
- 13% · 3
Top plugins in each topic
Ranked by trust score. A plugin is listed only when it has a score, no open advisory and no high-severity finding. How trust scores work.
Top 5 planning and workflow plugins
89 plugins in this topic, 80 eligible to rank.
- 1
Boss
echoVic · Trust score 95
BMAD pipeline plugin that orchestrates a full requirements-to-deploy workflow across nine specialist agents with an auditable runtime DAG and quality gates, for Claude Code…
- 2
dev-harness-kit
sanghee · Trust score 94
Enforced development workflow skills for Codex and Claude Code covering planning, TDD, debugging, review, security, CI, and release
- 3
claude-project-map
bokiko · Trust score 93
/project-map renders one status page: done / next / one decision box; scripts gather facts, the model judges
- 4
spec-superflow
MageByte · Trust score 93
Lean spec workflow with direct or planned execution, optional specifications, one final review, recoverable state, a bundled same-version CLI, Windows-compatible Qoder/WorkBuddy…
- 5
Ultra Mod
mertkayacs · Trust score 93
All-in-one mod pack for Claude Code: a usage limits and context HUD, a guard with undo for rm -rf and git reset --hard, .env and secret protection, and a receipt for every turn
Top 5 memory and context plugins
88 plugins in this topic, 76 eligible to rank.
- 1
Mnemoverse Memory
Mnemoverse · Trust score 96
Hosted AI agent memory that learns from outcomes, over remote MCP with OAuth sign-in: tell it a recalled memory helped or misled and it re-ranks what comes back next, with shared…
- 2
agent-kit
agent-kit-startup · Trust score 95
Extension pack of slash commands, skills, rules, and hooks adding plan-driven, confirmation-gated coding and staging-to-prod git workflows to Cursor and Claude Code
- 3
Magents
abnegate · Trust score 95
Shared session bus MCP + skill so Codex can list, read, hand off, send, and spawn with Claude, Cursor, Grok, and friends without pasting context.
- 4
Context Guard
Gerui Lv · Trust score 94
Preserves authoritative requirements and verification evidence across long-running Codex tasks and context compaction.
- 5
Salesforce Compound Engineering
Gella Sangamesh Gupta · Trust score 94
Salesforce-focused compound engineering plugin for Claude Code, Cursor, Codex, and other AI coding tools, with skills-first workflows, parallel persona dispatch, and Apex/LWC/Flow…
Top 5 code review plugins
77 plugins in this topic, 63 eligible to rank.
- 1
falsegreen-skill
Vinicius Queiroz · Trust score 95
Finds tests that stay green when the code they cover is broken, applying six ordered judgments over Python, TypeScript, JavaScript, and Robot Framework suites in Codex CLI and…
- 2
AtomLane
cloudguo123 · Trust score 94
Parallelizes only work proven safe by compiling resource-aware atomic plans for builds, tests, Make/Compose workflows, and long-running jobs with live progress and measured…
- 3
dev-harness-kit
sanghee · Trust score 94
Enforced development workflow skills for Codex and Claude Code covering planning, TDD, debugging, review, security, CI, and release
- 4
spec-superflow
MageByte · Trust score 93
Lean spec workflow with direct or planned execution, optional specifications, one final review, recoverable state, a bundled same-version CLI, Windows-compatible Qoder/WorkBuddy…
- 5
A Team
RBraga01 · Trust score 92
A complete engineering team in one folder with 26 specialist agents, 20 workflow skills, an orchestrator, and quality gates; integration and hook support vary by coding platform.
Top 5 docs and writing plugins
76 plugins in this topic, 64 eligible to rank.
- 1
HTML/CSS to Image API
HTML/CSS to Image · Trust score 95
Let AI agents capture live website screenshots, render HTML/CSS and populate reusable templates as images or PDFs without managing a browser
- 2
Long-form Distribution
Matthew Rowean · Trust score 95
Turns a published essay or newsletter issue into LinkedIn, X and Substack Notes posts cut from the author's own sentences, checks each post against the source, and tracks what…
- 3
bookstack-mcp
ttpears · Trust score 94
MCP server for searching, reading, and managing BookStack wiki content from AI assistants
- 4
humanizer-ru
Ilya Utov · Trust score 93
Agent skill for Claude Code, Codex, Cursor and Gemini that rewrites Russian text to remove 64 AI-generation markers (bureaucratese, calques, ChatGPT fingerprints), with a…
- 5
Zero Slop
Manav Mishra · Trust score 93
Say no to AI slop: a human-in-the-loop learning agentic workflow skill that scores text 0-100 for AI slop and rewrites it tastefully, with a standard-library Python scorer that…
Top 5 security and audit plugins
66 plugins in this topic, 51 eligible to rank.
- 1
Filesystem MCP
j0hanz · Trust score 95
MCP server: read, write, search, diff and patch files in allowed folders, with secret files blocked and confirmation before risky changes
- 2
Terraform Boundary
Somesh Ghaturle · Trust score 95
Claude Code, Codex, Cursor, and Windsurf skill that deploys AI agents on AWS, Azure, GCP, or Snowflake with Terraform while keeping a human in the loop: a hook blocks agent-run…
- 3
Salesforce Compound Engineering
Gella Sangamesh Gupta · Trust score 94
Salesforce-focused compound engineering plugin for Claude Code, Cursor, Codex, and other AI coding tools, with skills-first workflows, parallel persona dispatch, and Apex/LWC/Flow…
- 4
AW-1 Circuit Breaker
MvikManners · Trust score 93
Application-layer AST and argument policy gate for agent tool calls that checks configured rules before Python execution; complements, but does not replace, operating-system…
- 5
Ultra Mod
mertkayacs · Trust score 93
All-in-one mod pack for Claude Code: a usage limits and context HUD, a guard with undo for rm -rf and git reset --hard, .env and secret protection, and a receipt for every turn
Top 5 multi-agent orchestration plugins
57 plugins in this topic, 46 eligible to rank.
- 1
Boss
echoVic · Trust score 95
BMAD pipeline plugin that orchestrates a full requirements-to-deploy workflow across nine specialist agents with an auditable runtime DAG and quality gates, for Claude Code…
- 2
Herdr Codex Subagents
Konstantin Aristov · Trust score 94
Opens a read-only Herdr pane for every Codex subagent and closes it when the work is done.
- 3
A Team
RBraga01 · Trust score 92
A complete engineering team in one folder with 26 specialist agents, 20 workflow skills, an orchestrator, and quality gates; integration and hook support vary by coding platform.
- 4
Metis
Austin · Trust score 92
Repository-level engineering orchestrator that delegates bounded discovery, planning, implementation, review, and verification to fresh subagents, isolates mutable tasks in Git…
- 5
mstar-harness
Bohao Tang · Trust score 92
Multi-agent code harness plugin that routes work through PM, dev, QC, and QA roles with deterministic workflow gates enforced by a TypeScript engine, installable across dsh, omp…
Top 5 testing and tdd plugins
57 plugins in this topic, 48 eligible to rank.
- 1
falsegreen-skill
Vinicius Queiroz · Trust score 95
Finds tests that stay green when the code they cover is broken, applying six ordered judgments over Python, TypeScript, JavaScript, and Robot Framework suites in Codex CLI and…
- 2
Gremlyn
gremlyn-ai · Trust score 95
Security and chaos engineering for MCP agents: a Claude Code plugin that breaks your agent's MCP tools on purpose to test prompt-injection defenses and resilience
- 3
dev-harness-kit
sanghee · Trust score 94
Enforced development workflow skills for Codex and Claude Code covering planning, TDD, debugging, review, security, CI, and release
- 4
ThoughtProof MCP
ThoughtProof · Trust score 94
Local MCP pre-action verification for agents: mandate + proposed action → ALLOW/BLOCK/UNCERTAIN; execute only on ALLOW
- 5
Android Engineering Skills
Lennon Petrick · Trust score 93
Agent Skills for any skills-compatible coding agent (Claude Code, Codex, Gemini in Android Studio, and more) that teach how to structure, write, and test Android apps with Kotlin…
Top 5 images, audio and video plugins
55 plugins in this topic, 42 eligible to rank.
- 1
HTML/CSS to Image API
HTML/CSS to Image · Trust score 95
Let AI agents capture live website screenshots, render HTML/CSS and populate reusable templates as images or PDFs without managing a browser
- 2
Cadence Code
Michael Li · Trust score 93
Fully local voice conversations for Claude Code, Codex, Cursor, and Antigravity on Apple Silicon, with selectable MLX speech and transcription models
- 3
Mimi Seed
yoonion · Trust score 93
App launch ops for Claude Code and Codex: a local MCP server and CLI that drive Google Play, App Store Connect, Firebase, AdMob, and more from local credentials
- 4
homie
Prashanth K · Trust score 92
Four-level chat-voice personality layer for coding agents (off, yo, dawg, mafa) that changes how the agent talks in prose while leaving technical answers and code identical
- 5
WaveSpeed
WaveSpeedAI · Trust score 91
Generate and edit AI images, video, audio, and 3D on the WaveSpeed platform with a bundled MCP server and CLI skill, including live catalog search, local-file upload, and price…
Top 5 design and frontend plugins
46 plugins in this topic, 39 eligible to rank.
- 1
UXKIN
UXKIN · Trust score 93
Hosted MCP server and free skills that give Claude Code, Codex, Cursor and other agents real iOS app screens, user journeys and website design systems to reference before building…
- 2
EditLens
mymotomaker-ai · Trust score 92
EditLens highlights each Claude edit on the page you're building, with its file and line, and shows Claude's chat and code beside it
- 3
Frappe Agent
Dhairya Marwaha · Trust score 90
Frappe and ERPNext coding, customization, bench, and review intelligence for Codex.
- 4
Agentic Ship
moasq · Trust score 89
Cross-host product-development toolkit for Claude Code, Codex, Cursor, Hermes, and OpenClaw with shared rules, specialist roles, service connections, and machine-checked UI…
- 5
DataMagic
HKUSTDial · Trust score 88
A skill that teaches coding agents to plan and render narrated animated data videos from tabular data using DVSpec
Top 5 seo and marketing plugins
46 plugins in this topic, 41 eligible to rank.
- 1
Long-form Distribution
Matthew Rowean · Trust score 95
Turns a published essay or newsletter issue into LinkedIn, X and Substack Notes posts cut from the author's own sentences, checks each post against the source, and tracks what…
- 2
LVTD Skills
LVTD · Trust score 93
Reusable Agent Skills for Codex, Claude Code, and compatible clients, covering Django, Rust, Cookiecutter, SEO, traction, product marketing, and nonfiction publishing workflows.
- 3
Zero Slop
Manav Mishra · Trust score 93
Say no to AI slop: a human-in-the-loop learning agentic workflow skill that scores text 0-100 for AI slop and rewrites it tastefully, with a standard-library Python scorer that…
- 4
Agentic Ship
moasq · Trust score 89
Cross-host product-development toolkit for Claude Code, Codex, Cursor, Hermes, and OpenClaw with shared rules, specialist roles, service connections, and machine-checked UI…
- 5
GPT-6 Astra Outbound System
Pastel · Trust score 89
Codex plugin with 20 outbound skills and 12 agents that catch LinkedIn buying signals, qualify and rank leads, write human-sounding messages and posts, run sequences, and handle…
Top 5 devops and git plugins
45 plugins in this topic, 32 eligible to rank.
- 1
Terraform Boundary
Somesh Ghaturle · Trust score 95
Claude Code, Codex, Cursor, and Windsurf skill that deploys AI agents on AWS, Azure, GCP, or Snowflake with Terraform while keeping a human in the loop: a hook blocks agent-run…
- 2
Unified AI System
happy520ai · Trust score 94
Self-hosted AI gateway for Codex with provider-free prompt enhancement, fifteen governed MCP tools, and a credential-free Docker path.
- 3
Coolify
Severin Hilbert · Trust score 93
Control Coolify Cloud and self-hosted Coolify instances through API-aware workflow skills and local tools.
- 4
Infrawise
Sidd27 · Trust score 93
MCP server and CLI that gives AI coding assistants live AWS, database, and Kafka infrastructure context (table schemas, Lambda trigger event shapes, queue and IaC config) and…
- 5
DevRecap
João Quinteiro · Trust score 92
Reconstruct evidence-backed developer work from authorized Codex and Claude history plus read-only Git, then turn it into daily standups, weekly reviews, sprint recaps, and memory…
Top 5 research and search plugins
40 plugins in this topic, 31 eligible to rank.
- 1
Filesystem MCP
j0hanz · Trust score 95
MCP server: read, write, search, diff and patch files in allowed folders, with secret files blocked and confirmation before risky changes
- 2
Storyflo
Storyflo · Trust score 93
Agentic newsroom over a hosted MCP server with narrated briefings, a news-versus-prediction-market Divergence Index, and a searchable declassified archive.
- 3
Clawifi
ReARCLabs · Trust score 91
Claude Code plugin and MCP server for the hosted Clawifi API, giving agents web search, page fetch and scrape (including PDFs), site mapping, crawling and question-focused passage…
- 4
Exa Web Search
Codex Exa Community · Trust score 91
Search and fetch current web sources in Codex through the official Exa MCP server with browser OAuth.
- 5
Extensive Prompt Architect
Ext7 · Trust score 91
Turns rough ideas, briefs, and drafts into detailed, ready-to-use AI prompts for Claude Code and Codex
Top 5 debugging and observability plugins
39 plugins in this topic, 31 eligible to rank.
- 1
AtomLane
cloudguo123 · Trust score 94
Parallelizes only work proven safe by compiling resource-aware atomic plans for builds, tests, Make/Compose workflows, and long-running jobs with live progress and measured…
- 2
Herdr Codex Subagents
Konstantin Aristov · Trust score 94
Opens a read-only Herdr pane for every Codex subagent and closes it when the work is done.
- 3
SolidPing
Florent Clairambault · Trust score 93
Uptime monitoring from Codex through the hosted SolidPing MCP server with OAuth sign-in: read checks, results and incidents, diagnose failures, and manage checks and status pages…
- 4
spec-superflow
MageByte · Trust score 93
Lean spec workflow with direct or planned execution, optional specifications, one final review, recoverable state, a bundled same-version CLI, Windows-compatible Qoder/WorkBuddy…
- 5
Embedded Workbench
Amethyst Luna · Trust score 92
Embedded C/C++ firmware development toolbox — 7 skills (FreeRTOS, Keil MDK, ARMCLANG, HardFault triage, state machines, LVGL) plus workflow gates and 4 agents for Claude Code…
Top 5 payments and finance plugins
31 plugins in this topic, 24 eligible to rank.
- 1
Slashbooks
Gil Hildebrand · Trust score 94
Local-first AI bookkeeping plugin for Claude Cowork, Claude Code, and Codex that imports bank activity, closes the month, answers questions, and exports accountant-ready files for…
- 2
Ophis
Ophis · Trust score 93
Onchain token swaps for Codex via the hosted Ophis MCP server, MEV-protected and gasless, built on CoW Protocol.
- 3
PDFtoAgent (Dow Theory MCP)
babaksh · Trust score 92
Turnkey FastMCP server built with the PDFtoAgent pipeline, extracting 100+ deterministic tools from the book into a unified RAG + Skill + Tools bundle
- 4
AgentBadge MCP
AgentBadge · Trust score 91
Verify agents, check trust scores, and pay for services with x402 — ERC-8004 identity on Arc via MCP
- 5
BlindOracle
craigmbrown · Trust score 91
Joins a Grok Bot or Grok Build agent to the BlindOracle agent marketplace with a free ERC-8004 passport, role-scoped MCP tools, counterparty-risk controls for buying or selling…
Top 5 cost and token usage plugins
24 plugins in this topic, 22 eligible to rank.
- 1
AI Command Center
Hredo · Trust score 95
Local-first Windows, macOS and Linux command center for Claude Code, Codex, OpenCode, Aider, Gemini CLI, Ollama and 8,000+ API models, with live cost and token analytics (even for…
- 2
Codex Usage and Resets
Filament Labs · Trust score 92
Turns Codex usage into planning facts with linear pace, projected exhaustion, banked-reset expirations, and conservative unexpected-reset detection.
- 3
sectionsight
Jakub Dvořák · Trust score 91
Sets up cookieless, privacy-first website analytics (time per page section, consent banner, Slack or e-mail reports, health alerts) and proves the privacy claims with a…
- 4
Contexo
maheedhar132 · Trust score 87
Portable AI context and cost control across every AI coding harness
- 5
Encore Lite
Encore · Trust score 84
Free end-of-week surprise builder for Claude Code and Codex - the night before your weekly usage cap resets, it reads your recent work and builds one bonus deliverable, delivered…
How we counted
- We read every packaged plugin in the HOL catalog on . MCP servers and skills listed on their own are not included.
- Each topic is a fixed list of tags, for example code-review, pr-review and static-analysis for code review. Tags are lowercased and spaces become hyphens before matching.
- 437 plugins carry no tags and 207 carry only tags outside every topic. They count toward the catalog total and nowhere else.
- A topic appears once it holds at least 8 plugins.
- Runtime shares are declarations, not install tests. Check a listing before you install it in the agent you use.
Questions about plugin topics
What do most AI agent plugins do?
Planning and workflow is the largest topic: 89 of the 1,183 packaged plugins in the HOL catalog carry one of its tags, followed by memory and context with 88.
What is the highest-scored planning and workflow plugin?
Boss has a trust score of 95, the highest of the 80 planning and workflow plugins that have a score, no open advisory and no high-severity finding.
How are plugins assigned to topics?
Each topic is a fixed list of tags. A plugin joins every topic one of its tags names, so a plugin can sit in several topics and topic totals add up to more than the catalog. 437 plugins carry no tags and appear in no topic.
Does a high trust score mean a plugin is safe to install?
No. The score summarizes HOL’s automated checks of the listing and its source when the snapshot was taken. Read the listing, its permissions and its latest scan before you install anything.
Keep exploring
- Runtime compatibility map: which agent runtimes plugins support together.
- Plugin leaderboards: the plugins people install most, by runtime.
- Compare AI plugins: put two listings side by side before you choose.
- Browse AI plugins and MCP servers and search for the job you need done.