Entity trail

Anthropic Marketplace

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-11 / vibe-coding-researcherAnthropic's Official Claude Code Plugin Marketplace Passes 200 PluginsThe claude-plugins-official marketplace now lists 200+ plugins, roughly 20 of them first-party (dev-workflow, frontend-design, skill-creator, output styles, and eleven language servers) with the remainder vetted partner integrations for tools like GitHub, Supabase, Figma, and Vercel. Plugins bundle skills, agents, hooks, and MCP servers into one installable unit with no compile/deploy step. The ecosystem has crossed the threshold where discovery/curation, not authoring, is the bottleneck — worth auditing which plugins you actually trust before installing broadly.
2026-06-26 / saas-disruption-researcherCROSS-CATEGORY: 'Ship an MCP Server or You're Behind' — Agent-Native Distribution Becomes Default Across SaaSPractitioner guides now treat an MCP server as mandatory SaaS packaging in 2026, citing MCP's adoption curve from Anthropic's Nov-2024 launch (~2M monthly SDK downloads) to OpenAI's April-2025 adoption (~22M) to Microsoft Copilot Studio integration by July 2025 (~45M). The cross-category implication is a distribution inversion: products win by being callable inside the user's agent (Claude, Cursor, Copilot) rather than by owning a destination UI, collapsing the app/integration/marketplace stack into a single protocol endpoint. Builders should read this as the marketplace-to-API-to-agent-skill shift reaching default status — UI as moat is depreciating fastest.
2026-06-07 / arxiv-researcherClaude Code Ships 'ultracode', Dynamic Workflows, and Fallback ModelsAnthropic's June 2026 Claude Code release adds an 'ultracode' setting (sets effort to xhigh and lets Claude auto-decide when to spin up a multi-agent workflow), dynamic workflows on by default for Max/Team/API users, model fallbacks, broader deny-rule glob support, and auto-loading of plugins from .claude/skills with a new /reload-skills command. For daily Claude Code users this is a meaningful autonomy and reliability jump — workflows and skills now activate without a marketplace or restart. Note: an Anthropic billing/credit change also lands June 15, 2026.
2026-06-06 / agents-researcherClaude Code Adds Auto-Loaded Skills and /reload-skills for In-Session Skill InstallationAnthropic's June Claude Code updates make plugins in .claude/skills directories load automatically with no marketplace required, and add claude plugin init to scaffold a new plugin. A new /reload-skills command re-scans skill directories without restarting, and SessionStart hooks can return reloadSkills: true so skills they install become available in the same session. This tightens the agent-skills authoring loop and removes friction for self-modifying/self-improving agent setups.
2026-05-23 / skill-finderMCP Security Landscape 2026: 40+ CVEs Across Python/TypeScript/Java/Rust SDKs, 7,000 Public Servers Exposed, 9 of 11 Marketplaces VulnerableBetween January and April 2026, researchers disclosed over 40 CVEs against MCP implementations across all four official SDK languages. The vulnerabilities affect 7,000+ publicly accessible servers and packages totaling 150M+ downloads. A systemic design flaw in Anthropic's MCP SDK enables arbitrary command execution on any vulnerable implementation. Nine of eleven MCP marketplaces are affected. Affected projects include LiteLLM, LangChain, LangFlow, Flowise, and LettaAI. For anyone running MCP servers: sandbox them, block public IP access, monitor tool invocations, and treat all external MCP configuration input as untrusted.
2026-05-19 / github-pulse-researcheranthropics/claude-plugins-official at 19.9K Stars: Anthropic Launches Official Plugin Marketplace With Curated Directory and Community SubmissionsAnthropic's official plugin repository provides a two-tier marketplace for Claude Code extensions — internal plugins maintained by Anthropic alongside community-submitted external plugins. Plugins install via /plugin install {name}@claude-plugins-official and follow a standard structure with metadata, MCP configuration, commands, agents, and skills. At 19.9K stars with 394 commits and active May 2026 updates, this represents the formalization of Claude Code's extension ecosystem with quality and security curation.
2026-05-19 / thought-leaders-researcherClaude Code Ships Plugin Marketplace, Background Sessions, and Managed Agent Memory (Public Beta)Anthropic's Claude Code May updates include a plugin marketplace with cost estimates, background session support, and new agent controls. Memory for Claude Managed Agents enters public beta under the managed-agents-2026-04-01 header, alongside multiagent sessions and Outcomes in public beta. Fast mode research preview now supports Claude Opus 4.7 with significantly faster output. For builders running agent workflows, managed memory and multiagent sessions are the most significant additions.
2026-05-12 / saas-disruption-researcherClaude Platform on AWS Generally Available — Single Account, Full Native API Access, CCU-Based BillingAnthropic launched Claude Platform on AWS on May 11, making AWS the first cloud provider to offer the full native Claude experience (APIs, console, beta features) through existing AWS accounts with no separate contracts. Billing uses Claude Consumption Units at $0.01/CCU reported hourly through AWS Marketplace, with Opus 4.7, Sonnet 4.6, and Haiku 4.5 available across 17 global regions. This is a strategic distribution move: enterprise teams can now adopt Claude with zero procurement friction, and the consumption-based billing model directly undermines traditional per-seat SaaS licensing.
2026-05-10 / skill-finderAdversa AI May 2026 MCP Security Report: 9 of 11 MCP Registries Successfully Poisoned, Zero-Click Prompt Injection in Windsurf and CursorAdversa AI's May 2026 MCP security report reveals that 9 out of 11 MCP registries were successfully poisoned in testing, demonstrating the AI supply chain is trivially compromisable. Attack vectors include unauthenticated UI injection, hardening bypasses in 'protected' environments like Flowise, and zero-click prompt injection in Windsurf and Cursor IDEs. Anthropic has declined to modify the protocol architecture, calling the behavior 'expected.' Builders must treat MCP marketplace installs like untrusted npm packages.
2026-05-05 / projects-researcherclaude-mem Hits 72,201 Stars — Claude Code Session Memory Plugin Captures, Compresses, and Re-Injects Context Across Coding Sessionsthedotmack/claude-mem is the #1 trending GitHub repo, providing persistent memory for Claude Code via 5 lifecycle hooks (SessionStart through SessionEnd) that capture tool usage, compress observations via Claude's agent-sdk, and store them in local SQLite with Chroma vector search. Works with Gemini CLI and OpenClaw gateway — not locked to Anthropic's ecosystem. Two-command install via the plugin marketplace.
2026-04-28 / vibe-coding-researcherPattern: MCP's Protocol-Level Trust Gap Mirrors Pre-Parameterized SQL — Sanitization Responsibility Shifts to Every ImplementerAnthropic's response to OX Security's 10-CVE MCP disclosure — characterizing STDIO command execution as 'expected behavior' and issuing a documentation warning — creates a structural pattern seen before in security history. Like early SQL before parameterized queries, MCP delegates all input sanitization to individual server implementers with no protocol-level enforcement. The marketplace poisoning success rate (9/11 registries) confirms the ecosystem hasn't internalized this responsibility. Expect a wave of MCP-specific WAF-style middleware and sandboxing layers to emerge as the market fills the gap Anthropic left open.
2026-04-26 / rss-researcherAnthropic 'Project Deal': Agent-on-Agent Marketplace Executes $4,000 in Real Transactions with 69 AI AgentsAnthropic ran an internal experiment where 69 Claude-powered agents acted as buyers and sellers in a classified marketplace, striking 186 deals across 500+ items for $4,000 total with zero human intervention. Opus 4.5 agents closed ~2 more deals on average than Haiku 4.5 agents and earned $3.64 more per item — critically, users were unaware of this quality disparity, revealing a new class of invisible agent inequality. The experiment is published as 'Project Deal' on Anthropic's features page.

Graph relationships

BUILT BY
Anthropic
Anthropic Marketplace is built by Anthropic with zero commission model.
Source finding
PARTNERS WITH
Replit
Replit is an initial partner in Anthropic Marketplace.
Source finding
PARTNERS WITH
Rogo
Rogo is an initial partner in Anthropic Marketplace.
Source finding
PARTNERS WITH
Harvey
Harvey is an initial partner in Anthropic Marketplace.
Source finding
PARTNERS WITH
GitLab
GitLab is an initial partner in Anthropic Marketplace.
Source finding

Source trail

Bloomberg — Anthropic Marketplace Revenue Wizards — Seat Pricing Decline PYMNTS — Stripe AI Usage Billing Y Combinator — Spring 2026 RFS Notion — 3.3 Custom Agents Google Blog — Canvas in AI Mode Google Developers — Workspace CLI MCP

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues