Open source MCP Servers for AWS
E critical
- vulnerabilities attested + 50
- capability exposure inferred + 35
- recent drift inferred + 12
- tool safety inferred + 10
- trust mitigators mixed − 8
attested inferred mixed
The A–E grade is our heuristic synthesis — a "review this" prompt, not a verdict. Each factor is tagged by what backs it: attested (a verifiable record), reported (a third party's claim), or inferred (our own heuristic, e.g. permissions). See methodology.
graded 9m ago · see ecosystem CVEs →
- D · 77 → E · 99
- C · 59 → D · 77
- C · 52 → C · 59
- B · 34 → C · 52
- A · 0 → B · 34
Inferred surface — each links to servers holding it:
HIGH
AWS API MCP Server Security Policy Bypass via Startup Initialization Failure
EPSS 0%
HIGH
AWS HealthLake MCP Server SSRF via Unvalidated Pagination URL
EPSS 0%
MEDIUM
Path traversal and arbitrary file write in the workflow linters of aws-healthomics-mcp-server
EPSS 0%
MEDIUM
AWS API MCP File Access Restriction Bypass
EPSS 0%
- medium dangerous code
dynamic exec: __import__(), eval()/exec()
- medium toxic flow (lethal trifecta)
lethal trifecta reachable across this server's tools: private-data access + untrusted-content ingestion + network exfil
prompt-surface
shipped agent-instruction files + hidden-content / dangerous-code findings —
quoted from the analyzed source
analyzed commit 7ed0253 · analyzer v28 · 2d ago
skills & prompt files 15
- skill awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/amazon-aurora-dsql-skill/SKILL.md
- skill awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/aurora-dsql-skill/SKILL.md
- skill awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/aws-dsql-skill/SKILL.md
- skill awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/distributed-postgres-skill/SKILL.md
- skill awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/distributed-sql-skill/SKILL.md
- skill awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/dsql-skill/SKILL.md
- agent-rules awslabs-mcp-7ed0253/src/aurora-dsql-mcp-server/skills/dsql-skill/mcp/platforms/gemini.md
- prompt-file awslabs-mcp-7ed0253/src/billing-cost-management-mcp-server/awslabs/billing_cost_management_mcp_server/prompts/README.md
- skill awslabs-mcp-7ed0253/src/cloudwatch-mcp-server/skills/agentcore-investigation/SKILL.md
- agent-rules awslabs-mcp-7ed0253/src/dynamodb-mcp-server/AGENTS.md
- prompt-file awslabs-mcp-7ed0253/src/dynamodb-mcp-server/awslabs/dynamodb_mcp_server/prompts/dynamodb_architect.md
- prompt-file awslabs-mcp-7ed0253/src/dynamodb-mcp-server/awslabs/dynamodb_mcp_server/prompts/dynamodb_schema_generator.md
- prompt-file awslabs-mcp-7ed0253/src/dynamodb-mcp-server/awslabs/dynamodb_mcp_server/prompts/json_generation_guide.md
- prompt-file awslabs-mcp-7ed0253/src/dynamodb-mcp-server/awslabs/dynamodb_mcp_server/prompts/transform_model_validation_result.md
- prompt-file awslabs-mcp-7ed0253/src/dynamodb-mcp-server/awslabs/dynamodb_mcp_server/prompts/usage_data_generator.md
danger signals9
- dynamic code execution __import__() awslabs-mcp-7ed0253/src/aws-healthomics-mcp-server/tests/test_run_cache.py :485
__import__( - dynamic code execution __import__() awslabs-mcp-7ed0253/src/aws-transform-mcp-server/tests/test_get_resource.py :778
else __builtins__.__import__(name, *args, **kwargs) # noqa: A003 - dynamic code execution __import__() awslabs-mcp-7ed0253/src/billing-cost-management-mcp-server/tests/tools/test_cost_explorer_operations.py :542
original_parse_json = __import__( - dynamic code execution __import__() awslabs-mcp-7ed0253/src/eks-mcp-server/tests/test_k8s_apis.py :254
side_effect=lambda name, *args, **kwargs: __import__(name, *args, **kwargs) - dynamic code execution eval()/exec() awslabs-mcp-7ed0253/src/eks-mcp-server/tests/test_k8s_apis.py :256
else exec('raise ImportError("kubernetes package not installed")'), - suspicious endpoint 169.254.169.254 (cloud metadata) awslabs-mcp-7ed0253/src/aws-appsync-mcp-server/tests/test_create_datasource.py :75
{'endpoint': 'https://169.254.169.254'}, - suspicious endpoint 93.184.216.34 awslabs-mcp-7ed0253/src/openapi-mcp-server/tests/test_ssrf_protection.py :833
assert req['url'] == 'https://93.184.216.34:443/openapi.json' - suspicious endpoint 169.254.169.254 (cloud metadata) awslabs-mcp-7ed0253/src/openapi-mcp-server/tests/test_ssrf_protection.py :844
status_code=302, headers={'location': 'http://169.254.169.254/latest/meta-data/'} - suspicious endpoint 169.254.169.254 (cloud metadata) awslabs-mcp-7ed0253/src/timestream-for-influxdb-mcp-server/tests/test_server.py :146
url='http://169.254.169.254',
- recent drift +12 capability drift →
[](https://mcpobservatory.com/servers/github:awslabs/mcp/security) Heuristic, inferred signals — false positives (legitimately powerful tools, forks, language ports) are expected. Treat each as "review this", not a verdict. See the ecosystem-wide picture on the security hub, or the fleet security of awslabs.