# Vigil (npm · lv24-vigil)

Local-first security auditor: 263 deterministic rules, LLM triage, fixes and MCP tools

- Trust score: 84/100 (high trust)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-21

## Components

- npm · `lv24-vigil`: 84/100 (this document), [markdown](https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil.md), [page](https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil)

## Channel facts

- Registry: `npm`
- Package: `lv24-vigil`
- Version: `0.10.1`
- Transport: `stdio`

## Trust breakdown

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. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-11-21.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - No install/post-install scripts declared.
  - 0 of 6 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 100/100
  - Source repository is publicly reachable at the declared URL.
  - Cryptographically verified build provenance (signed, bound to LevapVeeskela/lv24-vigil).
  - Clear OSI-approved license (Apache-2.0).
  - Actively maintained (last published 2 days ago).
  - Publishes a security disclosure policy (SECURITY.md).
- **Schema Quality & AI Usability**: 82/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 2009 tokens (~118/item across 17 items; 14 tools + 3 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 20/100
  - Stability observed for 6 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 14 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 16 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 60/100
  - Spec-recency check failed: implements MCP spec 2025-06-18; the latest is 2026-07-28.

## Install

### How do I install the Vigil MCP server?

Vigil runs locally as an npm package, launched with npx -y lv24-vigil mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add levapveeskela-vigil -- npx -y lv24-vigil mcp
```

### Cursor

```json
{
  "mcpServers": {
    "levapveeskela-vigil": {
      "command": "npx",
      "args": [
        "-y",
        "lv24-vigil",
        "mcp"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "levapveeskela-vigil": {
      "command": "npx",
      "args": [
        "-y",
        "lv24-vigil",
        "mcp"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add levapveeskela-vigil -- npx -y lv24-vigil mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "levapveeskela-vigil": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "lv24-vigil",
        "mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add levapveeskela-vigil --command npx --arg -y --arg lv24-vigil --arg mcp
```

### Hermes

```yaml
mcp_servers:
  levapveeskela-vigil:
    command: "npx"
    args: ["-y", "lv24-vigil", "mcp"]
```

### Netclaw

```json
{
  "McpServers": {
    "levapveeskela-vigil": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "lv24-vigil",
        "mcp"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add levapveeskela-vigil -t stdio -c npx -a -y lv24-vigil mcp
```

### Other

```json
{
  "mcpServers": {
    "levapveeskela-vigil": {
      "command": "npx",
      "args": [
        "-y",
        "lv24-vigil",
        "mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-11-20 (score 84, +1)

No change was recorded against any check on this day. Stability & Change Management went from 13 to 17. That category is still filling its 30-day observation window: 4 days of observed history at the previous scan, 5 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-18 (score 83, +1)

No change was recorded against any check on this day. Stability & Change Management went from 7 to 10. That category is still filling its 30-day observation window: 2 days of observed history at the previous scan, 3 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-16 (score 82, +1)

- [security regression] Tool safety: pass → unverified
- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional regression] Capabilities: fail → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Schema quality: 100 → unverified
- [functional improvement] Stability: unverified → 0.03
- [functional] Package version: 0.9.3 → 0.10.1

### 2026-11-15 (score 81)

First indexed and scored.

## MCP tools (14)

### `vigil_scan` (~373 tokens)

Scan a repository with Vigil

Run Vigil's deterministic security rules over a repository or directory, optionally with model triage, deep review and a threat model. Returns a text summary and the full scan result as structuredContent. Scanned code is treated as data, never executed.

Input parameters:

- `baseRef` (string): Git ref to diff against when changedOnly is set (default HEAD).
- `changedOnly` (boolean): Scan only files changed against baseRef (requires a git repository).
- `deepReview` (boolean): Run the LLM deep review over high-risk files (requires a model).
- `diff` (string): Unified diff to scan instead of the working tree. Overrides changedOnly.
- `disableRules` (array): Skip these rule IDs.
- `exclude` (array): Glob patterns to exclude in addition to the project defaults.
- `include` (array): Glob patterns to include in addition to the project defaults.
- `maxFindings` (integer): Maximum findings to return (default 50, capped at 500).
- `minSeverity` (string): Minimum severity to report.
- `mode` (string): Scan depth. All modes run the deterministic rules; quick is the fastest LLM budget, standard is the default, deep adds an LLM deep review when a model is configured.
- `noBaseline` (boolean): Ignore the baseline and report suppressed findings as active.
- `onlyRules` (array): Run only these rule IDs (for example VGL-SEC-001).
- `path` (string): Directory or repository to scan. Defaults to the server's project root.
- `threatModel` (boolean): Also generate an LLM threat model and attach it to the result (requires a model).
- `useLlm` (boolean): Use the configured model. Omit to use the server default; false forces deterministic rules only.

### `vigil_review` (~148 tokens)

Review changed lines with Vigil

Scan only the added lines of a diff: the working-tree changes against a git ref, or a unified diff supplied by the client. Returns a text summary and the full scan result.

Input parameters:

- `baseRef` (string): Git ref to diff against (default HEAD).
- `changedOnly` (boolean): Review only changed files. Defaults to true; ignored when diff is supplied.
- `diff` (string): Unified diff to review. When omitted, changed files are read from git.
- `path` (string): Directory or repository to review. Defaults to the server's project root.
- `useLlm` (boolean): Use the configured model for triage. Omit to use the server default.

### `vigil_dependencies` (~181 tokens)

Audit dependency advisories with Vigil

Read dependency manifests and lockfiles, query the public OSV.dev database for the exact versions found and report known public advisories. Returns a text summary and the inventory, findings, coverage errors and limitations as structuredContent. Read-only: no package manager runs, no writes and no telemetry.

Input parameters:

- `includeDev` (boolean): Also query dev dependencies. Defaults to true.
- `maxPackages` (integer): Maximum distinct packages to query (default 200, capped at 1000).
- `osvBaseUrl` (string): Override the OSV API root. Defaults to https://api.osv.dev; must be https, or http on a loopback host (127.0.0.1, ::1, localhost). Credentials in the URL are refused.
- `path` (string): Directory or repository to audit. Defaults to the server's project root.

### `vigil_hardening` (~187 tokens)

Review anti-reverse-engineering hardening

Run only the anti-RE hardening rules (VGL-ARE-*): release minification, debug signing, debuggable builds, backup and cleartext settings, plist transport exceptions, release obfuscation, Electron/Node isolation, inspector flags and production source maps. Reports hardening gaps, not vulnerabilities. Optionally narrow the review to one platform. Returns a text summary grouped by platform and the full scan result as structuredContent.

Input parameters:

- `maxFindings` (integer): Maximum findings to return (default 50, capped at 500).
- `minSeverity` (string): Minimum severity to report.
- `noBaseline` (boolean): Ignore the baseline and report suppressed findings as active.
- `path` (string): Mobile or desktop project to review. Defaults to the server's project root.
- `platform` (string): Only report findings for this platform. Defaults to all.

### `vigil_mcp_audit` (~219 tokens)

Audit MCP server configuration

Run only the MCP configuration rules (VGL-MCP-*): unpinned server packages, literal secrets in env/headers/args, cleartext remote endpoints, disabled TLS verification, dangerous mounts, auto-approval and prompt-injection surfaces. Reads .mcp.json, mcp.json, mcp-config.json, claude_desktop_config.json, opencode.json[c], *.mcp.json and YAML variants; with includeUser it also reads known home-directory configs. Deterministic and offline: no configured server is started, no endpoint is contacted and nothing is written. Returns a text summary and the full audit result as structuredContent.

Input parameters:

- `includeUser` (boolean): Also read known user-level MCP configs under the home directory.
- `maxFindings` (integer): Maximum findings to return (default 50, capped at 500).
- `minSeverity` (string): Minimum severity to report.
- `path` (string): Directory to audit. Defaults to the server's project root; the resolved path must stay inside it.

### `vigil_explain` (~89 tokens)

Explain a Vigil finding

Explain one finding: what the rule detected, why it matters, an exploit scenario and the remediation direction. Uses the configured model when available; otherwise returns the deterministic rule description and remediation.

Input parameters:

- `findingId` (string, required): Finding ID (VGL-...), fingerprint (fp_...) or a unique prefix.
- `scanId` (string): Scan to look in. Defaults to the most recent scan.

### `vigil_fix` (~113 tokens)

Propose or apply a fix for a Vigil finding

Ask the configured model for a validated unified diff that fixes one finding. With apply=false (the default) nothing is written; with apply=true the diff is applied to the working tree and never committed.

Input parameters:

- `apply` (boolean): Apply the proposed diff to the working tree. Defaults to false (dry run).
- `findingId` (string, required): Finding ID (VGL-...), fingerprint (fp_...) or a unique prefix.
- `scanId` (string): Scan to look in. Defaults to the most recent scan.

### `vigil_threat_model` (~60 tokens)

Generate a threat model with Vigil

Ask the configured model for a repository threat model: assets, trust boundaries, entry points and the most likely attack paths. Requires a model.

Input parameters:

- `path` (string): Directory to model. Must match the server's project root; defaults to it.

### `vigil_status` (~41 tokens)

Show Vigil status

Report the engine version, project root, model configuration (with the key redacted), rule count, categories, store directory, the last scan summary and startup warnings.

### `vigil_rules` (~119 tokens)

List Vigil detection rules

List the deterministic detection rules with their severity, category, languages and title. Optionally filter by category, language, severity or a free-text query.

Input parameters:

- `category` (string): Only rules in this category.
- `language` (string): Only rules that apply to this language (for example go, python, javascript, yaml). Generic rules are included.
- `query` (string): Case-insensitive text matched against the ID, title, description, tags, CWE and OWASP fields.
- `severity` (string): Only rules with this base severity.

### `vigil_baseline` (~123 tokens)

Manage the Vigil suppression baseline

List, add, remove or clear accepted findings in the project baseline. Adding requires a finding ID and a reason; removed entries stop being suppressed on the next scan.

Input parameters:

- `action` (string): Baseline operation. Defaults to list.
- `findingId` (string): Finding ID (VGL-...), fingerprint (fp_...) or a unique prefix. Required for add and remove.
- `reason` (string): Why the finding is accepted. Required for add.
- `scanId` (string): Scan to resolve the finding in. Defaults to the most recent scan.

### `vigil_report` (~103 tokens)

Render a Vigil scan report

Render a stored scan as Markdown, JSON, SARIF or HTML. Without path the report is returned as text; with path it is written inside the app root and the absolute path is returned.

Input parameters:

- `format` (string): Report format. Defaults to md.
- `path` (string): Relative path inside the app root to write the report to (for example reports/vigil.md).
- `scanId` (string): Scan to render. Defaults to the most recent scan.

### `vigil_verify` (~108 tokens)

Verify that a Vigil finding is gone

Re-scan the finding's file (or explicit paths) with the deterministic rules and report whether the same fingerprint is still present. Useful after a fix.

Input parameters:

- `findingId` (string, required): Finding ID (VGL-...), fingerprint (fp_...) or a unique prefix.
- `paths` (array): Files to re-scan, relative to the scan root. Defaults to the finding's file.
- `scanId` (string): Scan the finding belongs to. Defaults to the most recent scan.

### `vigil_history` (~64 tokens)

List stored Vigil scans

List the most recent stored scans with their finding, critical and high counts. Scan IDs can be passed to vigil_explain, vigil_fix, vigil_report and vigil_verify.

Input parameters:

- `limit` (integer): Maximum number of scans to list (default 10, capped at 100).

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil#diagnostics

## Score history

- 2026-11-21: 84
- 2026-11-20: 84
- 2026-11-19: 83
- 2026-11-18: 83
- 2026-11-17: 82
- 2026-11-16: 82
- 2026-11-15: 81

## Common questions

### What is the Vigil MCP server?

Vigil is an MCP server listed in the public MCP registry as io.github.LevapVeeskela/vigil. Local-first security auditor: 263 deterministic rules, LLM triage, fixes and MCP tools. This page covers its npm package (lv24-vigil).

### Is the Vigil MCP server safe to use?

Vigil scores 84 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 21 November 2026. It declares no install or post-install scripts. 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 Vigil MCP server expose?

Vigil exposes 14 tools: vigil_scan, vigil_review, vigil_dependencies, vigil_hardening, vigil_mcp_audit, and 9 more. Their descriptions and schemas cost roughly 1,928 tokens of context every time the server is loaded.

### Is the Vigil MCP server still maintained?

Vigil is still listed as active in the MCP registry. We last reached this channel on 21 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 Vigil MCP server under?

Vigil declares the Apache-2.0 licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

### What did VerifyMCP test on the Vigil MCP server?

VerifyMCP ran 16 checks on Vigil, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.

## Links

- npm package: https://www.npmjs.com/package/lv24-vigil
- Socket report: https://socket.dev/npm/package/lv24-vigil
- Repository: https://github.com/LevapVeeskela/lv24-vigil
- Changelog RSS feed: https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil.xml
- Changelog JSON feed: https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil.json
- HTML version of this page: https://verifymcp.io/servers/levapveeskela-vigil/lv24-vigil
