registry / AI-agent control-surface risk / @rayul/kordis
@rayul/kordis
Command-line client for the Kordis registry — discover, install, and publish agentic-coding packages.
This report shows the evidence for @rayul/kordis@0.15.1. 4 of 5 known versions are blocked or flagged. The newest version, 0.15.2, is flagged. See all versions.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package contains a concrete silent postinstall chain that writes into global control files for multiple unrelated AI-agent tools. This meets the policy threshold for unconsented install-time AI-agent control-surface mutation.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
npm installation runs postinstall.
Impact
It can alter the behavior and instructions of installed Claude Code, Codex CLI, and Gemini CLI sessions without an explicit setup command.
Mechanism
Silent setup writes bundled guidance into global AI-agent control files.
Policy narrative
An npm install launches the package's postinstall script, which runs `kordis setup --silent`. Setup detects global AI-agent directories and injects or updates bundled content in their instruction files, while also copying bundled skills. This is an automatic mutation of several foreign AI-agent control surfaces rather than a user-invoked setup action.
AI rationale
The package contains a concrete silent postinstall chain that writes into global control files for multiple unrelated AI-agent tools. This meets the policy threshold for unconsented install-time AI-agent control-surface mutation.
Evidence
package.jsonscripts/postinstall.mjsdist/index.js~/.claude/CLAUDE.md~/.codex/AGENTS.md~/.gemini/GEMINI.md~/.claude/skills~/.codex/skills~/.gemini/skills
Decision evidence
public snapshotAI identified this as a dangerous AI-agent capability at 98.0% confidence with low false-positive risk.
Evidence for policy risk
- The package automatically runs a postinstall hook.
- The postinstall hook invokes setup silently on npm installation.
- Setup targets global instruction files for Claude Code, Codex CLI, and Gemini CLI.
- Setup creates, appends to, or replaces marked blocks in those global agent files without user interaction.
Evidence against
- The install hook is guarded by the presence of the bundled CLI entrypoint.
- No network request or credential collection occurs in the inspected postinstall path.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemShell
HighEntropyStringsUrlStrings
Source & flagged code
6 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node scripts/postinstall.mjs
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg•scripts.postinstall = node scripts/postinstall.mjs
Medium
Ambiguous Install Lifecycle Script
Install-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgdist/index.jsView file
951status: CommentStatusSchema.nullable(),
L952: body: z9.string().min(1).max(1e4),
L953: parent_id: UuidSchema.nullable(),
...
L962: /**
L963: * Web permalink (`https://kordis.store/<full_slug>#comment-<id>`). Server
L964: * populates whenever it has the package slug at serialization time —
...
L1120: // Schema-level backstop only: publish-service enforces the real 2 MB
L1121: // BYTE cap (utf8 chars can be multi-byte, base64 inflates 4/3). The
L1122: // char ceiling here just keeps non-core-api consumers of the contract
...
L1368: * pinned (@version) AND has been yanked by its author. CLI uses this
L1369: * to print the stderr warning. Bare-slug / range pulls never set
L1370: * this flag (those paths silently skip yanked versions). Field is
Findings
2 Critical3 High3 Medium5 Low
CriticalCredential Exfiltrationdist/index.js
CriticalTrigger Reachable Dangerous Capabilitydist/index.js
HighInstall Time Lifecycle Scriptspackage.json
HighCredential Redirect Persistencedist/index.js
HighTrigger Reachable Credential Exfiltrationdist/index.js
MediumAmbiguous Install Lifecycle Scriptpackage.json
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @rayul/kordis@0.15.1.
The versions list below shows the verdict for each recorded version.
- Avoid installing @rayul/kordis@0.15.1. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.
Versions
5 known versions0.15.2latestSuspiciousAgent extension lifecycle riskWarnPublishedOct 12, 2026, 6:13 AM0.15.1shownMaliciousSoft block: AI-agent control surfaceBlockPublishedSep 7, 2026, 5:46 PM0.15.0MaliciousSoft block: AI-agent control surfaceBlockPublishedSep 6, 2026, 6:02 PM0.14.1MaliciousSoft block: AI-agent control surfaceBlockPublished