agentsclimarketplace
Sign in

Shelves Context & AI engineering

Memory & context

2,410 rows from 996 repositories

What the agent knows between turns and between sessions.

What Memory & context skills agree on

807 skill files read, by 754 of the 1,056 authors on this shelf whose files we hold, 2026-09-06

The middle one of the 84 measured here is ~1.2k tokens long, counted with cl100k_base

Counted by distinct author, so one author publishing three of these counts once. Where a claim sits in fewer files than authors, that is said: a claim held by forty authors across three files is one file people copied, not forty people who agreed. Near-identical wordings are grouped and the other wordings are shown, so the grouping is yours to check.

What they tell the agent to do

  1. Preserve existing content structure15 of 754 in 9 filesalso worded as Preserve structure; Preserve existing structure when updating
  2. Front-load the leading word14 of 754 in 10 files
  3. Update existing entries instead of duplicating14 of 754 in 7 filesalso worded as Prefer updating existing keys over creating duplicates; Update existing entries instead of duplicates
  4. Keep CLAUDE.md under one hundred lines14 of 754 in 12 filesalso worded as Keep CLAUDE.md under 200 lines.; Keep root CLAUDE.md under 400 lines
  5. Read CLAUDE.md at the project root14 of 754also worded as Produce a CLAUDE.md at the project root; Write CLAUDE.md file to the project root
  6. Keep each meaning in a single source of truth12 of 754 in 8 files
  7. Redact sensitive information before committing11 of 754 in 4 filesalso worded as Redact sensitive details
  8. Scan for all CLAUDE.md files11 of 754 in 7 filesalso worded as Scan for all CLAUDE.md files in the repository; Find all CLAUDE.md files in the repository
  9. Use frontmatter for metadata on files10 of 754 in 3 filesalso worded as Use frontmatter for metadata
  10. Repeat user interactions 10 times10 of 754 in 4 filesalso worded as Repeat user interactions ten times
  11. Write the CLAUDE.md file into the target folder10 of 754 in 8 filesalso worded as Place CLAUDE.md files at the appropriate scope; Do not overwrite existing CLAUDE.md files
  12. Use memlab to process snapshots9 of 754 in 3 files
  13. Output the quality report before making any updates9 of 754 in 5 filesalso worded as Output the quality report before making updates; Output the quality report before making any updates
  14. Keep AGENTS.md under 60 lines9 of 754 in 7 filesalso worded as Keep AGENTS.md under 150 lines; Keep AGENTS.md under 200 lines
  15. Check every line for relevance8 of 754 in 4 filesalso worded as Check relevance line by line

What they tell it not to do

  1. Never store secrets or credentials30 of 754 in 27 filesalso worded as Never store secrets; Never store passwords or credentials
  2. Do not create duplicates16 of 754 in 9 filesalso worded as Do not create duplicate memories; Never create a near-duplicate
  3. Do not duplicate existing content14 of 754 in 12 filesalso worded as Do not duplicate skill content; Do not duplicate existing topics
  4. Never create AGENTS.md when CLAUDE.md exists13 of 754 in 12 filesalso worded as Do not populate CLAUDE.md with placeholder values; Never edit CLAUDE.md or AGENTS.md directly
  5. Do not overwrite existing configuration files entirely12 of 754also worded as Never overwrite an existing file silently; Do not overwrite a non-empty existing instruction file
  6. Do not let memory files grow unbounded11 of 754 in 3 files
  7. Do not read raw heapsnapshot files directly11 of 754 in 5 filesalso worded as Do not read raw heapsnapshot files; Do not attempt to read raw heapsnapshot files directly
  8. Do not update without user approval11 of 754 in 7 filesalso worded as Do not make updates without user approval; Do not save changes without user approval
  9. Do not include generic advice10 of 754also worded as Do not include generic best practices; Never generate generic advice
  10. Never edit past log entries9 of 754 in 6 filesalso worded as Never edit existing log entries; Don't edit existing entries

What they expect to be installed

  1. git116 of 754 in 111 files
  2. grep67 of 754
  3. python366 of 754 in 60 files
  4. bash53 of 754
  5. node44 of 754 in 41 files
  6. ls32 of 754 in 29 files
  7. python31 of 754
  8. find29 of 754 in 27 files
  9. gh26 of 754 in 25 files
  10. Glob20 of 754

What they ask it to produce

  1. CLAUDE.md132 of 754also worded as CLAUDE.md file; CLAUDE.local.md
  2. AGENTS.md90 of 754 in 89 filesalso worded as agents/<name>.md; AGENTS.override.md
  3. MEMORY.md27 of 754 in 25 filesalso worded as memory/glossary.md; memory/product.md
  4. Memory files23 of 754 in 18 filesalso worded as Updated memory files; Memory file
  5. SKILL.md19 of 754 in 17 filesalso worded as skill
  6. Search results17 of 754 in 11 filesalso worded as search results summary; memory search results
  7. Markdown files17 of 754 in 16 filesalso worded as Project markdown files; Memory markdown files
  8. README.md12 of 754also worded as wiki/README.md; docs/README.md

When Memory & context authors say to reach for one

The situations these authors wrote into their own files, counted out of the same 754 authors, with the skills that name each one

How Memory & context skills are built

2,132 skill directories by 835 authors, read from their repositories’ own file trees 2026-08-05

The middle bundle among those shipping files is 3 files, 14.1 KB beside SKILL.md

Counted by distinct author, same as above, so one author publishing forty template copies counts once. SKILL.md itself is not counted as a file, so a single-file skill is one where that file is the whole skill.

The shape

  1. SKILL.md is the whole skill339 of 835 authors, 981 of 2,132 skills
  2. files ship beside it496 of 835 authors, 1,151 of 2,132 skills
  3. executable scripts ship inside217 of 835 authors, 545 of 2,132 skills

The folders they converge on

  1. references/283 of 835 authors, 528 of 2,132 skills
  2. scripts/172 of 835 authors, 477 of 2,132 skills
  3. agents/91 of 835 authors, 135 of 2,132 skills
  4. assets/69 of 835 authors, 111 of 2,132 skills
  5. evals/44 of 835 authors, 68 of 2,132 skills
  6. templates/40 of 835 authors, 53 of 2,132 skills

  • Codebase Memory

    DeusData/codebase-memory-mcp/io.github.DeusData/codebase-memory-mcp MCP server

    39,060★ repo

    Codebase knowledge graph for AI agents — 159 languages, sub-ms queries, 99% fewer tokens.

  • Skillopt sleep

    microsoft/SkillOpt/plugins/claude-code/skills/skillopt-sleep Skill

    15,937★ repo

    SkillOpt is a text-space optimizer that trains reusable natural-language skills for frozen LLM agents through trajectory-driven edits, validation-gated updates, and deployable best_skill.md artifacts.

  • Sui Knowledge Base (kapa.ai)

    MystenLabs/sui/io.github.jessiemongeon1/sui-kapa MCP server

    7,735★ repo

    Semantic search across Sui documentation and knowledge sources, powered by kapa.ai

  • Basic memory

    basicmachines-co/basic-memory/io.github.basicmachines-co/basic-memory MCP server

    3,612★ repo

    Local-first knowledge management with bi-directional LLM sync via Markdown files.

  • Agent md refactor

    softaworks/agent-toolkit/dist/plugins/agent-md-refactor/skills/agent-md-refactor Skill

    2,311★ repo

    A curated collection of skills for AI coding agents. Skills are packaged instructions and scripts that extend agent capabilities across development, documentation, planning, and professional workflows.

  • Agents best practices

    DenisSergeevitch/agents-best-practices Skill

    2,219★

    Provider-neutral Agent Skill for Codex, Claude Code, and agentic harness design.

  • Memory

    memvid/claude-brain/skills/memory Skill

    547★ repo

    Give Claude Code photographic memory in ONE portable file. No database, no SQLite, no ChromaDB - just a single .mv2 file you can git commit, scp, or share. Native Rust core with sub-ms operations.

  • Ori memory

    aayoawoyemi/Ori-Mnemos/io.github.aayoawoyemi/ori-memory MCP server

    319★ repo

    Persistent memory infrastructure for AI agents. Identity, knowledge graph, and search.

  • Cognitive ai memory

    sachitrafa/cognitive-ai-memory/io.github.sachitrafa/cognitive-ai-memory MCP server

    no license263★ repo

    Ebbinghaus-based persistent memory for Claude. Memories decay with time, strengthen on recall.

  • Shodh memory

    varun29ankuS/shodh-memory/io.github.varun29ankuS/shodh-memory MCP server

    246★ repo

    Cognitive memory for AI agents — semantic search, Hebbian learning, knowledge graphs.

  • Neural memory

    nhadaututtheky/neural-memory/io.github.nhadaututtheky/neural-memory MCP server

    235★ repo

    Persistent memory for AI agents — 55 MCP tools, spreading activation recall.

  • OMEGA Memory

    omega-memory/omega-memory/io.github.omega-memory/omega-memory MCP server

    203★ repo

    Persistent memory, coordination, and learning for AI agents. Local-first, 25 MCP tools.

  • Harness engineering guide

    OdradekAI/harness-engineering-guide/harness-engineering-guide Skill

    143★ repo

    My personal directory of skills.

  • Agent memory implementation

    simbajigege/book2skills/skills/agent-memory-implementation Skill

    138★ repo

    Create best skills based on best books

  • Skill

    HelloRuru/claude-memory-engine/skill Skill

    132★ repo

    Claude Code 的記憶系統 | A memory system built with hooks + markdown. Zero dependencies.

  • Agentic readiness

    CodeAlive-AI/ai-driven-development/skills/agentic-readiness Skill

    122★ repo

    Practices, protocols, and skills for AI-driven software development. Skills and safety hooks for Claude Code, Codex, OpenCode, Cursor, Antigravity, and any agent supporting the Agent Skills standard.

  • Infranodus cli

    infranodus/skills/infranodus-cli Skill

    no license107★ repo

    Skills to augment LLM thinking process, integrated with InfraNodus insight generation tool

  • Agents md generator

    MADTeacher/mad-agents-skills/agents-md-generator Skill

    104★ repo

    Collection of agent skills for AI assistants working with Dart and Flutter projects

  • Neo4j agent memory skill

    neo4j-contrib/neo4j-skills/neo4j-agent-memory-skill Skill

    101★ repo

    Neo4j Skills for Coding and other Agents including Cypher

  • Memory mcp

    yuvalsuede/memory-mcp/io.github.yuvalsuede/memory-mcp MCP server

    98★ repo

    Persistent memory + git snapshots for Claude Code. Never lose context or code.

  • VelesDB Memory

    cyberlife-coder/VelesDB/io.github.cyberlife-coder/velesdb-memory MCP server

    no license87★ repo

    Offline agentic memory: remember/recall/relate/forget/why over a fused vector+graph+columnar engine

  • Agent compat sync

    wquguru/skills/skills/agent-compat-sync Skill

    no license85★ repo

    Practical Agent Skills — English-for-engineers coaching, Pi Agent setup, and more. Install via npx skills add.

  • Audit context

    DheerG/swarms/skills/audit-context Skill

    84★ repo

    Achieve extraordinary results with claude code across a variety of tasks

  • Agents context management

    PaulRBerg/agent-skills/skills/agents-context-management Skill

    69★ repo

    PRB's collection of agent skills

  • Ham

    kromahlusenii-ops/ham Skill

    67★

    (HAM) Memory system for AI coding agents. Cut token usage by 80% by scoping context to directories.

  • Creating claude md

    sruthik27/creating-claude-md Skill

    63★

    Claude Code / Codex skill: generate a lean, high-signal CLAUDE.md from scratch via repo scan + focused user interview.

  • Wrap

    ViryaZheng/wrap-skill/plugins/wrap/skills/wrap Skill

    61★ repo

    A Claude Code skill that consolidates scattered project markdown files into a single canonical CLAUDE.md, with a pinned PROGRESS block at the top.

  • Synap Memory

    maximem-ai/maximem_synap_sdk/ai.maximem/synap MCP server

    60★ repo

    Persistent memory for AI agents — log and recall conversation context over MCP.

  • Memory Vault

    MihaiBuilds/memory-vault/io.github.MihaiBuilds/memory-vault MCP server

    60★ repo

    Local-first AI memory layer with hybrid search. Postgres + pgvector. Self-hosted, MIT.

  • emem, the verifiable memory protocol for the physical world

    Vortx-AI/emem/io.github.Vortx-AI/emem MCP server

    52★ repo

    Shared, verifiable memory for AI agents and robots: signed tokens that resolve and verify offline.

  • T2a memory

    t2a-protocol/T2A/integrations/t2a-memory/templates/skills/t2a-memory Skill

    48★ repo

    Tool-to-Agent Protocol: tools can be smart without embedded LLM calls.

  • Perseus Vault (formerly Mimir/Mneme), persistent memory engine for AI agents

    Perseus-Computing-LLC/perseus-vault/io.github.Perseus-Computing-LLC/perseus-vault MCP server

    47★ repo

    Persistent, deterministic memory for AI agents. Local-first: SQLite, FTS5, AES-256-GCM, 55+ tools.

  • lean-memory

    Wuesteon/lean-memory/io.github.Wuesteon/lean-memory MCP server

    47★ repo

    Embedded, local-first agent memory: one SQLite file per namespace, no server, no cloud.

  • Code memory

    kapillamba4/code-memory/io.github.kapillamba4/code-memory MCP server

    47★ repo

    Local semantic code search with Git history. Works offline, no API key needed.

  • Agents md manager

    tdimino/claude-code-minoan/skills/core-development/agents-md-manager Skill

    41★ repo

    A curated ~/.claude/ configuration for professional development workflows — 90+ skills, 46 hooks, and CLI tools

  • Ai memory

    alphaonedev/ai-memory-mcp/io.github.alphaonedev/ai-memory MCP server

    41★ repo

    Persistent memory for any AI — zero token cost until recall

  • Fewword

    sheeki03/Few-Word/plugins/fewword/skills/fewword Skill

    no license38★ repo

    Claude Code plugin that offloads large outputs to filesystem and retrieves when required.

  • Claude code project

    zebbern/skills/claude-code-project Skill

    no license29★ repo

    High-quality Agent Skills made to perform reliably in real workflows

  • Cortex Memory Engine

    gambletan/cortex/io.github.gambletan/cortex MCP server

    29★ repo

    Local-first AI memory engine — 4-tier memory, people graph, Bayesian beliefs. Encrypted, 62µs.

  • MCP Memory Gateway

    IgorGanapolsky/mcp-memory-gateway/io.github.IgorGanapolsky/mcp-memory-gateway MCP server

    26★ repo

    Pre-action gates that block AI agents from repeating known mistakes.

  • Agent instructions

    nimadorostkar/Claude-Skills-collection/skills/agent-tooling/agent-instructions Skill

    25★ repo

    A curated library of 137 production-grade skills for Claude and other AI coding agents.

  • hmem — Humanlike Memory for AI Agents

    Bumblebiber/hmem/io.github.Bumblebiber/hmem-mcp MCP server

    24★ repo

    Persistent 5-level hierarchical memory for AI agents. SQLite-backed, lazy-loaded.

  • SAME - Stateless Agent Memory Engine

    sgx-labs/statelessagent/io.github.sgx-labs/same MCP server

    no license23★ repo

    Persistent memory for AI coding agents. Local-first SQLite + vector search. 17 MCP tools.

  • Skill memory

    memodb-io/skill-memory/skills/skill-memory Skill

    no license21★ repo

    agent skill as memory layer

  • M3 Memory

    skynetcmd/m3-memory/io.github.skynetcmd/m3-memory MCP server

    21★ repo

    Local-first agentic memory — 100+ tools, 92% LongMemEval-S, hybrid search, GDPR, zero cloud.

  • Memory custodian

    waittim/MemoryCustodian/skills/memory-custodian Skill

    20★ repo

    Durable, repo-native project memory for coding agents—without context bloat.

  • Genesys memory

    Astrix-Labs/genesys/io.github.Astrix-Labs/genesys-memory MCP server

    20★ repo

    Open-source causal memory for AI agents: persistent, explainable, MCP-native memory (13 tools).

  • Neverinfamous memory journal mcp

    neverinfamous/memory-journal-mcp/ai.smithery/neverinfamous-memory-journal-mcp MCP server

    19★ repo

    A MCP server built for developers enabling Git based project management with project and personal…

  • Compaction advisor

    vignesh07/compaction-advisor Skill

    17★

    Automatic context monitoring for Claude Code — never get interrupted by mid-task compaction again.

  • Mnemoverse Memory

    mnemoverse/mcp-memory-server/io.github.mnemoverse/mcp-memory-server MCP server

    17★ repo

    Hosted memory for AI agents that learns and forgets — one key across Claude, Cursor & ChatGPT.

  • Lore

    TheaDust/lore Skill

    16★

    This is an AI coding agent Skill that provides framework-agnostic project memory, enabling agents to preserve architecture, decisions, and conventions across sessions.

  • Ruledoctor

    syf2211/ruledoctor/plugin/skills/ruledoctor Skill

    15★ repo

    Agent Skill: read CLAUDE.md & project rules before coding. Optional CLI audit + hooks. Claude/Codex/Cursor.

  • Ui design memory

    oneworks-ai/app/.oo/skills/ui-design-memory Skill

    15★ repo

    One Works application monorepo for AI agents, plugins, desktop, web, CLI, and Relay.

  • Pmll memory mcp

    drQedwards/PPM/io.github.drQedwards/pmll-memory-mcp MCP server

    15★ repo

    PMLL Memory MCP — persistent KV context memory and Q-promise deduplication.

  • Neurodivergent memory

    jmeyer1980/neurodivergent-memory/io.github.jmeyer1980/neurodivergent-memory MCP server

    15★ repo

    Persistent knowledge graph MCP server for neurodivergent thinking. BM25 search, no cloud LLM.

  • Remembra — AI Memory Layer

    remembra-ai/remembra/io.github.remembra-ai/remembra MCP server

    15★ repo

    Persistent AI memory with entity resolution, temporal knowledge graph & 11 MCP tools.

  • Linksee memory

    michielinksee/linksee-memory/io.github.michielinksee/linksee-memory MCP server

    14★ repo

    Catches when your project drifts from its own decisions. Local-first cross-agent memory MCP.

  • Synapse Layer — Zero-Knowledge Memory

    SynapseLayer/synapse-layer/io.github.SynapseLayer/synapse-secure-memory MCP server

    13★ repo

    Zero-Knowledge Context™ & Neural Handover™ for AI Agents. The Trust Layer for AI Memory.

  • Global Agent Memory

    ozankasikci/global-agent-memory/io.github.ozankasikci/global-agent-memory MCP server

    13★ repo

    Shared, project-aware memory for Claude Code, Codex, and other agents with human review.

  • Create agents md

    MSiccDev/ai-context-kit/skills/create-agents-md Skill

    12★ repo

    AI Context Kit provides a structured, instruction-based framework for context-aware AI collaboration across LLM providers. It includes authoritative specs, canonical templates, and skills workflows for creating and validating user context and project AGENTS.md files.

All 2,410 in the catalog