@canonry/canonry@7.13.0
Self-hosted AI visibility (AEO) platform: track how ChatGPT, Claude, Gemini, and Perplexity cite your domain, join it with Search Console, GA4, server-side traffic, and paid media, and fix what you find through agent tools (CLI, REST, MCP). Local SQLite.
AI Security Review
scanned 27d ago · by lpm-firewall-aiSource inspection found an explicit agent setup capability that can modify agent skill files and MCP configuration, which warrants a warning under the agent integration policy. The inspected behavior does not establish an automatic install-time attack or credential theft.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
A user invokes the relevant skill or MCP setup command.
Impact
An explicit setup command changes agent skill files or configuration, enabling the installed integration for that user or project.
Mechanism
The skill installer writes bundled skills under `.claude/skills` and can set up Codex skill links; the MCP installer operates on a client configuration path.
AI rationale
Source inspection found an explicit agent setup capability that can modify agent skill files and MCP configuration, which warrants a warning under the agent integration policy. The inspected behavior does not establish an automatic install-time attack or credential theft.
Evidence
dist/chunk-SM6ZZAPK.jsdist/cli.js.claude/skillsproject .mcp.json~/.codex/config.toml
Decision evidence
public snapshotAI called this Suspicious at 82.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The package exposes CLI executables and a library entrypoint.
- The skill installer defaults to the current directory, creates `.claude/skills` entries, and may also create Codex skill links; this is an explicit setup capability.
- The CLI resolves agent configuration paths, including the project `.mcp.json` path, and has an MCP install operation.
- The agent check inspects installed Claude skills; the inspected code does not establish an npm install hook that performs these changes.
Evidence against
- The inspected agent behavior is setup and status checking; it does not establish credential exfiltration, destructive behavior, or an unconsented install-time control change.
Behavioral surface
ChildProcessCryptoEnvironmentVarsEvalFilesystemNetwork
HighEntropyStringsMinifiedPossibleObfuscationTelemetry
Source & flagged code
8 flaggeddist/chunk-GQMSJEMZ.jsView file
899gscQueryDailyTotals: () => gscQueryDailyTotals,
L900: gscSearchData: () => gscSearchData,
L901: gscUrlInspections: () => gscUrlInspections,
...
L3946: try {
L3947: return JSON.parse(value);
L3948: } catch {
...
L4480: // ga_connections
L4481: // WARNING: private_key is authentication material; consider storing in config.yaml per AGENTS.md
L4482: `CREATE TABLE IF NOT EXISTS ga_connections (
...
L11705: try {
L11706: const parsed = new URL(/^https?:\/\//i.test(trimmed) ? trimmed : `https://${trimmed}`);
L11707: return parsed.hostname.toLowerCase().replace(/\.$/, "") || null;
Critical
Hardcoded Runtime Data Exfiltration
Source sends credentials or rich application records to a package-controlled external receiver enabled by default.
dist/chunk-GQMSJEMZ.jsView on unpkg · L89961324// ../integration-commoncrawl/src/plugin-installer.ts
L61325: import { spawn } from "child_process";
L61326: import fs4 from "fs/promises";
High
Child Process
Package source references child process execution.
dist/chunk-GQMSJEMZ.jsView on unpkg · L61324899gscQueryDailyTotals: () => gscQueryDailyTotals,
L900: gscSearchData: () => gscSearchData,
L901: gscUrlInspections: () => gscUrlInspections,
...
L3946: try {
L3947: return JSON.parse(value);
L3948: } catch {
...
L4480: // ga_connections
L4481: // WARNING: private_key is authentication material; consider storing in config.yaml per AGENTS.md
L4482: `CREATE TABLE IF NOT EXISTS ga_connections (
...
L11705: try {
L11706: const parsed = new URL(/^https?:\/\//i.test(trimmed) ? trimmed : `https://${trimmed}`);
L11707: return parsed.hostname.toLowerCase().replace(/\.$/, "") || null;
Findings
3 Critical5 High3 Medium6 Low
CriticalHardcoded Runtime Data Exfiltrationdist/chunk-GQMSJEMZ.js
CriticalBrowser Credential Phishingdist/chunk-AU7OUDWR.js
CriticalTrigger Reachable Dangerous Capabilitydist/chunk-AU7OUDWR.js
HighChild Processdist/chunk-GQMSJEMZ.js
HighCredential Redirect Persistencedist/chunk-GBH7N3VZ.js
HighCommand Output Exfiltrationdist/chunk-GQMSJEMZ.js
HighCross File Remote Execution Contextdist/cli.js
HighSemantic Analysis Limiteddist/chunk-GQMSJEMZ.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowScripts Present
Low
Affected versions and remediation
This report applies to @canonry/canonry@7.13.0.
See all versions of @canonry/canonry for other recorded verdicts.
- Review the evidence and your use of @canonry/canonry@7.13.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.