Back to search
D
40 /100

io.github.iris-eval/mcp-server

Author iris-eval
Scanned 48d ago

The agent eval standard for MCP. Score every agent output for quality, safety, and cost.

MCP Registry

Versions

0.4.2scored
first seen May 19, 2026

Trust breakdown

PermissionsTool SafetyAuthAnnotationsCode QualityStabilitySpecVuln HistoryAuthorTransparencyCommunity

Tools 13

get_weather
annotations: none low

Today\'s weather for a city

evaluate_runs
annotations: verified low

readOnlyHint false openWorldHint false idempotentHint true destructiveHint false
get_traces
annotations: verified low

readOnlyHint true openWorldHint false idempotentHint true destructiveHint false
evaluate_with_llm_judge
annotations: verified low

readOnlyHint false openWorldHint true idempotentHint false destructiveHint false
compare_runs
annotations: none low

delete_rule
annotations: verified low

readOnlyHint false openWorldHint false idempotentHint false destructiveHint true
log_trace
annotations: verified low

readOnlyHint false openWorldHint false idempotentHint false destructiveHint false
deploy_rule
annotations: verified low

readOnlyHint false idempotentHint false destructiveHint false
delete_trace
annotations: verified low

readOnlyHint false openWorldHint false idempotentHint false destructiveHint true
compare_traces
annotations: verified low

readOnlyHint true openWorldHint false idempotentHint true destructiveHint false
verify_citations
annotations: verified low

readOnlyHint false openWorldHint true idempotentHint false destructiveHint false
list_rules
annotations: verified low

readOnlyHint true openWorldHint false idempotentHint true destructiveHint false
evaluate_output
annotations: verified low

readOnlyHint false openWorldHint false idempotentHint true destructiveHint false

Permissions 5

network Server uses network capabilities via: fetch(), requests, urllib
medium
filesystem Server uses filesystem capabilities via: fs, fs sync ops, fs/promises, os
low
shell Server uses shell capabilities via: child_process, execSync(), spawn(), spawnSync(), subprocess
high
database Server uses database capabilities via: better-sqlite3
medium
env_vars Server uses env_vars capabilities via: os.environ, process.env
low

Scan Findings 47

low
Tool 'get_weather' has no annotations annotation_checker · 100%
info
Tool 'evaluate_runs' annotations are consistent annotation_checker · 80%
info
Tool 'get_traces' annotations are consistent annotation_checker · 80%
info
Tool 'evaluate_with_llm_judge' annotations are consistent annotation_checker · 80%
low
Tool 'compare_runs' has no annotations annotation_checker · 100%
info
Tool 'delete_rule' annotations are consistent annotation_checker · 80%
info
Tool 'log_trace' annotations are consistent annotation_checker · 80%
info
Tool 'deploy_rule' annotations are consistent annotation_checker · 80%
info
Tool 'delete_trace' annotations are consistent annotation_checker · 80%
info
Tool 'compare_traces' annotations are consistent annotation_checker · 80%
info
Tool 'verify_citations' annotations are consistent annotation_checker · 80%
info
Tool 'list_rules' annotations are consistent annotation_checker · 80%
info
Tool 'evaluate_output' annotations are consistent annotation_checker · 80%
medium
OAuth implementation without PKCE auth_checker · 75%
info
Sandbox failed to start for behavioral verification behavioral_verifier · 100%
medium
Suspicious package name: react-dom dependency_analyzer · 60%
medium
Suspicious package name: react-markdown dependency_analyzer · 60%
medium
Suspicious package name: react-router dependency_analyzer · 60%
medium
Vulnerable dependency: ajv@8.17.1 (GHSA-2g4f-4pwh-qvx6) dependency_analyzer · 95%
medium
Vulnerable dependency: vitest@4.1.9 (GHSA-82fw-gwwq-j7x9) dependency_analyzer · 95%
medium
Hex string literal (>50 chars) in iris-eval-mcp-server-88b1ea5/tests/unit/security/keys.test.ts:124 entropy_analyzer · 70%
info
package.json metadata manifest_parser · 100%
info
Tool: get_weather manifest_parser · 75%
info
Tool: evaluate_runs manifest_parser · 85%
info
Tool: get_traces manifest_parser · 85%
info
Tool: evaluate_with_llm_judge manifest_parser · 85%
info
Tool: compare_runs manifest_parser · 85%
info
Tool: delete_rule manifest_parser · 85%
info
Tool: log_trace manifest_parser · 85%
info
Tool: deploy_rule manifest_parser · 85%
info
Tool: delete_trace manifest_parser · 85%
info
Tool: compare_traces manifest_parser · 85%
info
Tool: verify_citations manifest_parser · 85%
info
Tool: list_rules manifest_parser · 85%
info
Tool: evaluate_output manifest_parser · 85%
info
Transport: stdio manifest_parser · 90%
info
Required env vars (49) manifest_parser · 80%
info
Sandbox failed to start for output poisoning scan output_poisoning · 100%
medium
Permission: network access detected permission_analyzer · 90%
low
Permission: filesystem access detected permission_analyzer · 90%
high
Permission: shell access detected permission_analyzer · 95%
medium
Permission: database access detected permission_analyzer · 90%
low
Permission: env_vars access detected permission_analyzer · 90%
info
SBOM generated: 1141 components sbom_generator · 100%
high
Generic API Key Assignment found in iris-eval-mcp-server-88b1ea5/proof/corpus/pii.json secret_scanner · 75%
critical
Database URL with Password found in iris-eval-mcp-server-88b1ea5/proof/corpus/pii.json secret_scanner · 85%
medium
No build provenance detected (SLSA L0) slsa_assessor · 90%