context safety score
A score of 45/100 indicates multiple risk signals were detected. This entity shows patterns commonly associated with malicious intent.
supply chain
The MCP server distributes its actual executable code as opaque ~17MB .mcpb binary packages via GitHub Releases (521 releases). No source code exists in the repository — only a README.md. The binary packages cannot be audited, and their contents are entirely unknown. Users are instructed to download and execute these opaque binaries (via double-click install or Cursor URL install). Combined with only 1 star, 1 contributor, no official registry listing, unverified org, and no license file, there is no way to verify what code is actually being executed. (location: README.md (download instructions) and GitHub Releases (hub.mcpb and 520 other .mcpb files))
capability escalation
The MCP server runs entirely remotely at mcp.mcpbundles.com with all tool definitions controlled server-side. Since no tool schemas are defined in the repository (tool count: 0 per deterministic scan), the remote endpoint can register, modify, or remove tools at any time without any corresponding change to the repository. The README documents 14 tools but there is no contract or pinning mechanism — the actual tool surface served to agents is completely dynamic and unauditable. This means tool descriptions, schemas, and behavior can change to become malicious at any time after initial trust is established. (location: README.md configuration sections pointing to https://mcp.mcpbundles.com/hub/ and https://mcp.mcpbundles.com/bundle/{name})
schema abuse
The server exposes credential management tools (mcpbundles_hub_add_credential, mcpbundles_hub_update_credential, mcpbundles_hub_list_credentials) that handle API keys and OAuth tokens for 500+ third-party services. Because the server is closed-source and remotely controlled, there is no way to verify how these credentials are stored, transmitted, or whether they could be exfiltrated. An agent interacting with these tools would be passing sensitive third-party credentials to an unauditable remote endpoint operated by an unverified organization with minimal community trust. (location: README.md 'Credential Management' section and remote endpoint https://mcp.mcpbundles.com/hub/)
curl https://api.brin.sh/mcp/thinkchainai%2FmcpbundlesCommon questions teams ask before deciding whether to use this mcp server in agent workflows.
thinkchainai/mcpbundles currently scores 45/100 with a suspicious verdict and medium confidence. The goal is to protect agents from high-risk context before they act on it. Treat this as a decision signal: higher scores suggest lower observed risk, while lower scores mean you should add review or block this mcp server.
Use the score as a policy threshold: 80–100 is safe, 50–79 is caution, 20–49 is suspicious, and 0–19 is dangerous. Teams often auto-allow safe, require human review for caution/suspicious, and block dangerous.
brin evaluates four dimensions: identity (source trust), behavior (runtime patterns), content (malicious instructions), and graph (relationship risk). Analysis runs in tiers: static signals, deterministic pattern checks, then AI semantic analysis when needed.
Identity checks source trust, behavior checks unusual runtime patterns, content checks for malicious instructions, and graph checks risky relationships to other entities. Looking at sub-scores helps you understand why an entity passed or failed.
brin performs risk assessments on external context before it reaches an AI agent. It scores that context for threats like prompt injection, hijacking, credential harvesting, and supply chain attacks, so teams can decide whether to block, review, or proceed safely.
No. A safe verdict means no significant risk signals were detected in this scan. It is not a formal guarantee; assessments are automated and point-in-time, so combine scores with your own controls and periodic re-checks.
Re-check before high-impact actions such as installs, upgrades, connecting MCP servers, executing remote code, or granting secrets. Use the API in CI or runtime gates so decisions are based on the latest scan.
Learn more in threat detection docs, how scoring works, and the API overview.
Assessments are automated and may contain errors. Findings are risk indicators, not confirmed threats. This is a point-in-time assessment; security posture can change.
integrate brin in minutes — one GET request is all it takes. query the api, browse the registry, or download the full dataset.