AI agent tools for prompt engineering
In this catalog, a prompt engineering tool is one that writes, tests or manages prompts and agent instructions. Our enrichment pass tags 173 listings with it: 18 MCP servers, 85 agent skills, 60 plugins, 4 agent frameworks, 5 AI agents and 1 no-code agent builder. By listing rank, the top MCP server is Packmind, the top agent skill is caveman-compress and the top plugin is Nexpath. 10 of the 173 are published by the vendor behind the product, and 95 were active in the last 30 days. 2 of the 18 MCP servers run only as remote servers, with nothing to install locally. Each entry below links to its listing page, with grades, install commands and sources.
Updated Oct 26, 2026 · Built from catalog data; how we rank and grade
What the job is
- Tagged listings
- 173
- By type
- 18 MCP servers, 85 agent skills, 60 plugins, 4 agent frameworks, 5 AI agents and 1 no-code agent builder
- Published by the vendor
- 10 of 173
- Active in the last 30 days
- 95 of 173
- Open source
- 147 of 173
- MCP servers that run only remotely
- 2 of 18
- Hosts named by the MCP servers
- Claude Code (18), Cursor (18), VS Code (18)
Building blocks by type
The top entries of each type, by listing rank. Open a listing for its install command, grades and sources.
MCP servers 18 tagged
- 1Packmind
packmind.aiOfficial
Packmind captures, scales, and enforces your organization's technical decisions.
- Published by the vendor
- Remote server, no local install
- Active in the last 30 days
★ 319Last active Oct 22, 2026Remote (HTTP)Claude Code, VS Code, Cursor, Windsurf
AFreshness: A–Security: not graded yet - 2Skillmd
skillmd.comOfficial
Search, fetch, lint, and install Agent Skills (SKILL.md) from the SkillMD registry.
- Published by the vendor
- Runs locally or as a remote server
- Active in the last 30 days
★ 190Last active Oct 9, 2026Local (stdio), Remote (HTTP)Claude Code, VS Code, Codex, Cursor +2
AFreshness: A–Security: not graded yet - 3Claude Prompts MCP
minipuft
Reusable prompt templates, multi-step workflow chains, and quality gates for any MCP client.
- Active in the last 30 days
- Open source (MIT)
★ 189Last active Oct 24, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2
AFreshness: A–Security: not graded yet - 4Vibe Check MCP Server
PV-Bhat
Metacognitive AI agent oversight: adaptive CPI interrupts for alignment, reflection and safety
- Most GitHub stars here (502)
- Open source (MIT)
★ 502Last active Aug 20, 2026Local (stdio)Claude Code, Codex, VS Code, Cursor +2
BFreshness: B–Security: not graded yet - 5SuperCompress
Supercompress
Compress coding-agent context ~64%.
- Runs locally or as a remote server
- Active in the last 30 days
- Open source (MIT)
★ 137Last active Oct 22, 2026Local (stdio), Remote (HTTP)Claude Code, VS Code, Codex, Cursor +2
AFreshness: A–Security: not graded yet
Agent skills 85 tagged
- 1caveman-compress
JuliusBrussee
Compress a memory file such as CLAUDE.md or a todo list into caveman format to save input tokens, keeping a readable backup.
- Largest repository here (114,786 stars)
- Active in the last 30 days
- Open source (Apache-2.0)
★ 114.8KLast active Oct 17, 2026Claude Code, Claude apps, Codex, Cursor +6
AFreshness: A–Security: not graded yet - 2agents-generator
OJPalenzuela
Generate project-specific AGENTS.md and companion rules by analyzing a codebase.
- Repository has 48.4K stars
- Active in the last 30 days
- Open source (MIT)
★ 48.4KLast active Oct 24, 2026Claude Code
AFreshness: A–Security: not graded yet - 3configuration-hardening
langchain-aiOfficial
Review Talon tool placement and defensive prompts when requested or when adding or changing tools or subagents; apply confirmed changes and verify active…
- Published by the vendor
- Repository has 30.6K stars
- Active in the last 30 days
★ 30.6KLast active Oct 23, 2026Claude Code, Claude apps, Codex, Cursor +6
AFreshness: A–Security: not graded yet - 4skill-creator
PrimeIntellect-ai
Create, validate, and install Prime Agent skills - both markdown skills and Python-backed skills callable from the Python kernel.
- Repository has 22.3K stars
- Active in the last 30 days
★ 22.3KLast active Oct 21, 2026Claude Code, Claude apps, Codex, Cursor +6
AFreshness: A–Security: not graded yet - 5context-optimization
muratcankoylan
This skill should be used for improving context efficiency: context budgeting, observation masking, prefix or KV-cache strategy, partitioning, token-cost…
- Repository has 18.2K stars
- Active in the last 30 days
- Open source (MIT)
★ 18.2KLast active Oct 1, 2026Claude Code, Claude apps, Codex, Cursor +6
AFreshness: A–Security: not graded yet
Plugins 60 tagged
- 1
Nexpathnexpath
Strengthen your prompts and catch skipped checks before shipping.
- Active in the last 30 days
- Open source (Apache-2.0)
★ 358Last active Oct 24, 2026VS Code, Cursor, Windsurf, VSCodium
AFreshness: A–Security: not graded yet - 2
BAML v0Boundary
BAML v0 is a DSL for AI applications.
- Most GitHub stars here (9,431)
- Active in the last 30 days
- Open source (Apache-2.0)
★ 9.4KLast active Oct 24, 2026VS Code, Cursor, Windsurf, VSCodium
AFreshness: A–Security: not graded yet - 3
Prompt Driven DevelopmentPromptDriven
PDD CLI installation and syntax highlighting for .prompt files.
- Active in the last 30 days
- Open source (MIT)
★ 913Last active Sep 28, 2026VS Code, Cursor, Windsurf, VSCodium
AFreshness: A–Security: not graded yet - 4
Live LLM Token Counterbedirt
Live token counts and token-boundary highlighting for GPT, Claude, Gemini, and HuggingFace tokenizers.
- Open source (MIT)
★ 73Last active Aug 30, 2026VS Code, Cursor, Windsurf, VSCodium
BFreshness: B–Security: not graded yet
Agent frameworks 4 tagged
- 1Ax
Ax
DSPy-style toolkit for typed LLM signatures, agents and flows, TypeScript-first with Python, Java, C++, Go and Rust libraries.
- 3K GitHub stars
- Active in the last 30 days
- Open source (Apache-2.0)
★ 3K↓ 27.2K/wkLast active Oct 21, 2026
AFreshness: A–Security: not graded yet - 2DSPy
Stanford NLP
Python framework for writing language-model programs as code and optimizing their prompts and weights instead of hand-tuning prompts.
- Most GitHub stars here (38,766)
- Most weekly downloads here (1.3M)
- Active in the last 30 days
★ 38.8K↓ 1.3M/wkLast active Oct 22, 2026
AFreshness: A–Security: not graded yet - 3guidance
guidance-ai
Python library for steering language models by constraining output with regex and grammars and mixing control flow with generation.
- 21.8K GitHub stars
- Open source (MIT)
- For Python
★ 21.8K↓ 3.2K/wkLast active May 21, 2026
CFreshness: C–Security: not graded yet - 4AgentForge
DataBassGit
Low-code Python framework for YAML-defined agents and multi-agent Cogs, with personas, prompt templates and memory nodes.
- Open source (GPL-3.0)
- For Python
★ 860↓ 241/wkLast active Aug 10, 2026
BFreshness: B–Security: not graded yet
Search agent frameworks for prompt engineering → All harnesses & sdks
AI agents 5 tagged
- 1Bring Your AI
bringyour.ai
Paid local-first AI agent setup toolkit: Claude Code to Codex migration, central Claude/Codex/Cursor sync, distill snapshots, first-setup builds, no-data…
CFreshness: C–Security: not graded yet - 2CommonRig
Benys Labs
An agent operating kit for Claude Code and Codex: single-sourced rules, enforcement hooks, a file-based work ledger, protocols, generic skills, a setup prompt…
CFreshness: C–Security: not graded yet - 3StudioMeyer Crew
StudioMeyer
Agent Personas for any LLM client that speaks MCP.
CFreshness: C–Security: not graded yet - 4Agent Therapy
Bencium
Independent counselling for AI agents that are stuck: retrying the same failing step, facing conflicting instructions, unable to decide, pressured by a user…
CFreshness: C–Security: not graded yet - 5Marz Greta-Lock Network
Marz Greta-Lock Network
Add provisional emotional-interaction guidance to an agent you already operate across human-facing roles.
CFreshness: C–Security: not graded yet
No-code builders 1 tagged
- 1Rivet
Ironclad
Open-source visual IDE for building AI agents and prompt chains as graphs that run inside your own application.
- 4.7K GitHub stars
- Open source (MIT)
★ 4.7KLast active Aug 26, 2026
BFreshness: B–Security: not graded yet
Search no-code builders for prompt engineering → All no-code builders
Questions
- What are the top tools for prompt engineering?
- By listing rank, the top MCP server is Packmind, the top agent skill is caveman-compress and the top plugin is Nexpath. Rank weighs adoption, freshness, documentation and publisher identity; money never changes it.
- How many tools does the catalog list for prompt engineering?
- 173 listings are tagged for it: 18 MCP servers, 85 agent skills, 60 plugins, 4 agent frameworks, 5 AI agents and 1 no-code agent builder. Tags come from an automated enrichment pass over the top-ranked listings, so tools outside that pass are not counted.
- Which AI apps do the MCP servers work with?
- Of the 18 MCP servers, 18 list Claude Code, 18 list Cursor and 18 list VS Code as a supported host. Open a listing for its install command and the other hosts it names.
- Which of these are published by the vendor itself?
- 10 of the 173 are published by the vendor behind the product, including configuration-hardening, skill-creator and discernment-nudge.
173 listings are tagged for this use case; the lists above show the top entries.