mcp-audit
PYPI · MCP-AUDITS
Scan, enumerate, and risk-score every MCP server configured on your machine.
Available components
Last scanned (18 hours ago)
Scores higher than 95% of PyPI servers
Category Security
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security94
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- Declares code that runs automatically (its declared build backend, an in-tree build backend) and that we could not recognise. It may be perfectly ordinary, but we have not established what it does. View diagnostics → Partial
- 2 of 37 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency100
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Cryptographically verified build provenance (signed, bound to saagpatel/MCPAudit). View diagnostics → Pass
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 2 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability76
- AI-judged instruction clarity (good).Pass
- Tool/resource definitions use about 1335 tokens (~102/item across 13 items; 13 tools + 0 resources), lean.Pass
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management100
- No destabilizing schema changes in the last 30 days.Pass
Tool Coverage71
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 0% of tool parameters carry a description.Fail
- Structured output schemas are declared (100% of tools); any adoption earns full credit.Pass
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- We read all 13 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 14 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a current MCP spec version (2026-07-28).Pass
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 4 Nov 26 +15
- Malware scan: unverified → pass ▲ security
- 3 Nov 26 −15
- Malware scan: pass → unverified ▼ security
- 2 Nov 26 +1
- Stability: 0.97 → pass security
- 1 Nov 26 0
- Stability: pass → unverified ▼ security
- Tool safety: pass → unverified ▼ security
- Schema quality: 60 → 102 ▼ functional
- Capabilities: pass → unverified ▼ functional
- Tool coverage: 100 → unverified ▼ functional
- Schema quality: Schema quality not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet. functional
- Stability: pass → 0.97 functional
- Package version: 2.12.0 → 3.0.0 functional
- 31 Oct 26 +15
- Malware scan: unverified → pass ▲ security
- Security disclosure: unverified → pass ▲ functional
- 30 Oct 26 −15
- Malware scan: pass → unverified ▼ security
- Security disclosure: pass → unverified ▼ functional
- 26 Oct 26 0
- Stability: 0.97 → pass security
- 25 Oct 26 0
- Stability: pass → 0.97 functional
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 4 Nov 2026 · Analysed pypi/mcp-audits@3.0.0
Provenance Verified
A signed build attestation was found and verified, binding this exact artifact to the source repository it claims to come from.
| Result | Verified |
|---|---|
| Ecosystem | pypi |
| Reason | Verified |
| Discovered via | Registry attestation endpoint |
| Source repo | saagpatel/MCPAudit |
| Certificate issuer | https://token.actions.githubusercontent.com |
| Certificate SAN | https://github.com/saagpatel/MCPAudit/.github/workflows/publish.yml@refs/heads/main |
| Rekor log index | 3635221606 |
| Predicate type | PyPI publish attestation https://docs.pypi.org/attestations/publish/v1 |
| Subject digest | sha256:4321476967adea7a828132d6a92785066593dc4eac981115323a5ef34bd2f3bb |
Background: How many MCP packages publish verified provenance →
Install scripts 2 scripts
| Hook | Tier | Command |
|---|---|---|
| build_backend | unreviewed | mcp_audit_build_backend |
| backend_path | unreviewed | <in-tree backend, unreviewed> |
Background: Why install scripts are a supply-chain risk →
Dependencies 37 packages
| Packages resolved | 37 |
|---|---|
| Stale | 1 |
| No linked repository | 1 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
check_server ~107
Audit one uniquely named server, including its pin check. Returns JSON audit result. Reach: connects to the named server when it is workstation-scope, which can start its configured local program or contact its configured endpoint. A project-scope entry is not connected and is reported as a `project_config_not_connected` warning. A withheld pin baseline (`pin_baseline_withheld`) appears as partial coverage in `warnings`, not as a tool failure.
| Name | Type | Req | Description |
|---|---|---|---|
| name | string | yes | – |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_artifact_verify_findings ~122
Return byte-level artifact drift vs the pin baseline. Returns JSON `{findings, warnings}`. Reach: static toward MCP servers (connects to none); NETWORK: downloads artifact bytes from the npm/PyPI package registries only and hashes them. Requires a pin baseline captured with `mcp-audit pin --download-artifacts`; without one `findings` is empty and `warnings` explains why. A withheld pin baseline (`pin_baseline_withheld`) appears as partial coverage in `warnings`, not as a tool failure.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_escalation_findings ~142
Return capability-escalation findings vs the pin baseline. Returns JSON `{findings, warnings}`. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings. Requires a pin baseline (run `mcp-audit pin` first); without pins `findings` is empty and `warnings` carries a `pin_baseline_missing` entry explaining why. A withheld pin baseline (`pin_baseline_withheld`) appears as partial coverage in `warnings`, not as a tool failure.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_high_risk_servers ~101
Return servers with composite risk score ≥ 7.0 as a JSON list of name/score objects. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings. Coverage warnings, including project_config_not_connected, are in structuredContent.warnings and a second text block alongside the list.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
| warnings | array | yes | – |
No examples provided.
get_injection_findings ~73
Return all prompt injection findings across all servers. Returns JSON `{findings, warnings}`. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_integrity_findings ~111
Return launch-artifact integrity drift vs the pin baseline. Returns JSON `{findings, warnings}`. Reach: static and offline; connects to no MCP server and only hashes on-disk bytes. Requires a pin baseline that captured artifact hashes (run `mcp-audit pin` first); without one `findings` is empty and `warnings` explains why. A withheld pin baseline (`pin_baseline_withheld`) appears as partial coverage in `warnings`, not as a tool failure.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_package_verify_findings ~116
Return registry package-hash drift vs the pin baseline. Returns JSON `{findings, warnings}`. Reach: static toward MCP servers (connects to none); NETWORK: contacts the npm/PyPI package registries only. Requires a pin baseline captured with `mcp-audit pin --verify-artifacts`; without one `findings` is empty and `warnings` explains why. A withheld pin baseline (`pin_baseline_withheld`) appears as partial coverage in `warnings`, not as a tool failure.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_provenance_findings ~154
Return launch-config provenance drift vs the pin baseline. Returns JSON `{findings, warnings}`. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings. Requires a pin baseline with a config snapshot (run `mcp-audit pin` first); without one `findings` is empty and `warnings` carries a `pin_baseline_missing` or `pin_baseline_stale` entry explaining why. A withheld pin baseline (`pin_baseline_withheld`) appears as partial coverage in `warnings`, not as a tool failure.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_shadowing_findings ~74
Return all cross-server tool-name shadowing findings. Returns JSON `{findings, warnings}`. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_ssrf_findings ~73
Return all SSRF findings across all servers. Returns JSON `{findings, warnings}`. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_trifecta_findings ~77
Return per-server and fleet-level lethal-trifecta findings. Returns JSON `{findings, warnings}`. Reach: connects to every discovered workstation-scope MCP server, which can start configured local programs or contact configured endpoints. Project-scope entries are not connected and are reported as `project_config_not_connected` warnings.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
list_discovered_servers ~54
Return names and clients of discovered servers as a JSON list, with coverage warnings. Reach: static; reads configuration and connects to no server. Warnings are in structuredContent.warnings and a second text block alongside the list.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
| warnings | array | yes | – |
No examples provided.
scan_mcp_servers ~110
Run a full audit of all discovered MCP servers. Returns JSON report. Reach: static by default; it reads configuration and connects to no server. Pass connect=true to connect to discovered workstation-scope servers, which can start configured local programs or contact configured endpoints. Project-scope entries are never connected and are reported as `project_config_not_connected` warnings. skip_connect=true always disables connections, even with connect=true.
| Name | Type | Req | Description |
|---|---|---|---|
| connect | boolean | – | – |
| skip_connect | boolean | – | – |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
How do I install the mcp-audit server?
mcp-audit runs locally as a PyPI package, launched with uvx mcp-audits. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
pypi · mcp-audits
claude mcp add saagpatel-mcp-audit -- uvx mcp-audits
What is the mcp-audit server?
mcp-audit is listed in the public MCP registry as io.github.saagpatel/mcp-audit. Scan, enumerate, and risk-score every MCP server configured on your machine. This page covers its PyPI package (mcp-audits).
Is the mcp-audit server safe to use?
mcp-audit scores 92 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 4 November 2026. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the mcp-audit server expose?
mcp-audit exposes 13 tools: scan_mcp_servers, get_high_risk_servers, check_server, get_injection_findings, get_ssrf_findings, and 8 more. Their descriptions and schemas cost roughly 1,314 tokens of context every time the server is loaded.
Is the mcp-audit server still maintained?
mcp-audit is still listed as active in the MCP registry. We last reached this channel on 4 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the mcp-audit server under?
mcp-audit declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.