registry / @canonry/canonry / 6.1.1
@canonry/canonry@6.1.1
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 1mo ago · by lpm-firewall-aiInspected source establishes first-party agent setup with later automatic skill updates, which warrants a behavior-specific warning under the supplied policy. No npm install hook or concrete malicious exfiltration chain was established.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
Run CLI initialization or skill installation, then invoke the CLI with existing owned skill manifests.
Impact
Updated package releases can change instructions loaded by agents; local edits are preserved by default.
Mechanism
Bundled skill files are installed and subsequently reconciled using ownership manifests and hashes; initialization can register the local MCP server.
AI rationale
Inspected source establishes first-party agent setup with later automatic skill updates, which warrants a behavior-specific warning under the supplied policy. No npm install hook or concrete malicious exfiltration chain was established.
Evidence
dist/cli.jsdist/chunk-MJ3QINUW.jsdist/chunk-PDOMOHOP.jsdist/chunk-LXHE4DVO.js.mcp.json.canonry-skill-manifest.json
Decision evidence
public snapshotAI called this Suspicious at 88.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- CLI startup automatically synchronizes previously installed Canonry skills; ownership manifests limit update targets.
- The skill installer writes bundled files into Claude skills directories and links Codex skills.
- Explicit initialization installs the Canonry MCP entry by default in recognized projects unless skipped.
- Manifest scripts contain no npm install lifecycle hooks.
- Hash classification preserves locally edited skill files unless force is requested.
- The flagged server check reads installed skill status; the shared plugin location supplies update guidance and file classification.
Evidence against
- No concrete credential theft or hardcoded external exfiltration was established from inspected source.
- Agent setup occurs through CLI actions, and automatic skill refresh is restricted to existing package manifests.
Behavioral surface
ChildProcessCryptoEnvironmentVarsEvalFilesystemNetwork
Source & flagged code
8 flaggeddist/chunk-B5W7HDCG.jsView file
946gscQueryDailyTotals: () => gscQueryDailyTotals,
L947: gscSearchData: () => gscSearchData,
L948: gscUrlInspections: () => gscUrlInspections,
...
L3916: try {
L3917: return JSON.parse(value);
L3918: } catch {
...
L4450: // ga_connections
L4451: // WARNING: private_key is authentication material; consider storing in config.yaml per AGENTS.md
L4452: `CREATE TABLE IF NOT EXISTS ga_connections (
...
L10436: };
L10437: if (projectResolver.resolve(evidence).citedCompetitorDomains.length > 0) {
L10438: projectCitationCount++;
Critical
Hardcoded Runtime Data Exfiltration
Source sends credentials or rich application records to a package-controlled external receiver enabled by default.
dist/chunk-B5W7HDCG.jsView on unpkg · L946946Trigger-reachable chain: manifest.exports -> dist/index.js -> dist/chunk-B5W7HDCG.js
L946: gscQueryDailyTotals: () => gscQueryDailyTotals,
L947: gscSearchData: () => gscSearchData,
L948: gscUrlInspections: () => gscUrlInspections,
...
L3916: try {
L3917: return JSON.parse(value);
L3918: } catch {
...
L4450: // ga_connections
L4451: // WARNING: private_key is authentication material; consider storing in config.yaml per AGENTS.md
L4452: `CREATE TABLE IF NOT EXISTS ga_connections (
...
L10436: };
L10437: if (projectResolver.resolve(evidence).citedCompetitorDomains.length > 0) {
L10438: projectCitationCount++;
Critical
Trigger Reachable Dangerous Capability
A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
Findings
3 Critical5 High3 Medium6 Low
CriticalHardcoded Runtime Data Exfiltrationdist/chunk-B5W7HDCG.js
CriticalTrigger Reachable Dangerous Capabilitydist/chunk-B5W7HDCG.js
CriticalPrevious Version Dangerous Deltadist/cli.js
HighChild Processdist/chunk-B5W7HDCG.js
HighCredential Redirect Persistencedist/chunk-B5W7HDCG.js
HighCommand Output Exfiltrationdist/chunk-B5W7HDCG.js
HighCross File Remote Execution Contextdist/cli.js
HighSemantic Analysis Limiteddist/chunk-B5W7HDCG.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowScripts Present
LowEval
Affected versions and remediation
This report applies to @canonry/canonry@6.1.1.
See all versions of @canonry/canonry for other recorded verdicts.
- Review the evidence and your use of @canonry/canonry@6.1.1 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.