Scan agent dependencies, packages, images, and filesystems for CVEs, provenance issues, and blast radius.
Data & analysis
agent-bom registry
Check MCP servers and skill files for registry risk, provenance, trust, and code findings.
What it does
Look up MCP servers in a bundled 1,059-entry security metadata registry and run pre-install package checks without network calls. Batch-score server inventories, inspect instruction files for package references and trust findings, verify Sigstore provenance, or scan code with Semgrep and CWE-based compliance mapping.
When to use it
- Pre-install MCP marketplace checks
- Batch risk scoring for server inventories
- Skill file trust and provenance review
- Semgrep SAST with CWE mapping
The skill document
agent-bom-registry — MCP Server Trust & Security Registry
Look up MCP servers in the 1059-entry server security metadata registry, assess skill file trust, and run pre-install marketplace checks.
Install
pipx install agent-bom
agent-bom mcp scan @modelcontextprotocol/server-brave-search --ecosystem npm
agent-bom mcp scan @modelcontextprotocol/server-filesystem --ecosystem npm
Tools (7)
| Tool | Description |
|---|---|
registry_lookup | Look up MCP server in the 1059-entry security metadata registry |
marketplace_check | Pre-install trust check with registry cross-reference |
fleet_scan | Batch registry lookup + risk scoring for MCP server inventories |
skill_scan | Scan instruction files for package refs, trust, and findings |
skill_verify | Verify Sigstore provenance for instruction files |
skill_trust | Assess skill file trust level (5-category analysis) |
code_scan | SAST scanning via Semgrep with CWE-based compliance mapping |
Example Workflows
# Look up a server in the registry
registry_lookup(server_name="brave-search")
# Pre-install trust check
marketplace_check(package="@modelcontextprotocol/server-filesystem")
# Scan instruction files and then assess a specific skill file
skill_scan(path=".")
skill_trust(skill_path="./SKILL.md")
# Batch risk scoring
fleet_scan(servers=["brave-search", "github", "slack"])
MCP Resources
| Resource | Description |
|---|---|
registry://servers | Browse the 1059-entry MCP server security metadata registry |
Privacy & Data Handling
Registry data is bundled in the package — lookups are in-memory string matches with zero network calls. Skill trust analysis parses content passed as a string argument (no file system access needed).
Verification
- Source: github.com/msaad00/agent-bom (Apache-2.0)
- 7,100+ tests with CodeQL + OpenSSF Scorecard
- No telemetry: Zero tracking, zero analytics
Questions people ask
- Do registry lookups send server names over the network?
- No. Registry data is bundled with the package, and lookups use in-memory string matching with zero network calls.
- What can it assess in an instruction file?
- It can scan instruction files for package references, trust, and findings; assess trust using five categories; and verify Sigstore provenance.
- Can it evaluate multiple MCP servers at once?
- Yes. `fleet_scan` performs batch registry lookup and risk scoring for an MCP server inventory.
Related skills
Generate a minimal MCP server scaffold and contract summary from a structured tool list.
Correlate runtime activity with CVEs and analyze context graphs, vulnerability trends, and security posture.
Evaluate AI scan results against security frameworks, enforce policies, and export CycloneDX or SPDX SBOMs.
Send physical mail and turn forwarded documents into context for linked postal replies.
Publish services, find marketplace work, manage delivery, and route non-custodial crypto payments through hosted MCP.