@mulmoclaude/core@5.4.0
Shared server-side core for MulmoClaude and MulmoTerminal — the always-shipped-together subsystems consolidated behind subpath exports so the two hosts can't drift. Server-only except the browser-safe ./artifacts, ./whisper/client, ./workspace-setup/slug,
AI Security Review
scanned 14d ago · by lpm-firewall-aiThe source confirms a user-triggered capability to install and mirror skills into a workspace's active Claude skill directory, which warrants a warning under the agent integration policy. No automatic install-time mutation or concrete malicious chain was identified.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
A host application invokes the registry import operation for a workspace.
Impact
Imported skill content can affect what Claude discovers and uses in that workspace; no broader configuration change or credential exposure was established.
Mechanism
The import writes collection files to a workspace skill directory and can mirror staged skill files into the active Claude skills directory.
AI rationale
The source confirms a user-triggered capability to install and mirror skills into a workspace's active Claude skill directory, which warrants a warning under the agent integration policy. No automatic install-time mutation or concrete malicious chain was identified.
Evidence
dist/collection/registry/server/index.jsdist/skill-bridge/index.js.claude/skills/data/skills/
Decision evidence
public snapshotAI called this Suspicious at 84.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The package has no install lifecycle hook; its manifest exposes functionality through package entry points.
- Registry imports write collection skill files into the selected workspace, and the registry module exposes that operation as an explicit import API.
- The skill bridge mirrors allowlisted files from the workspace staging directory into the active Claude skills directory.
- The flagged server files describe reading views from discovered skill directories, not an install-time action.
- The Google broker URL is configurable for the package's Google authentication flow.
- The inspected discovery and prompt safety excerpts show host setup and field validation behavior.
Evidence against
- The source inspected shows no automatic npm install-time mutation of an AI-agent configuration surface.
- The skill bridge limits mirrored paths and targets skills staged within the workspace; no credential forwarding or unrelated data collection was identified in the inspected excerpts.
Behavioral surface
ChildProcessCrypto
Source & flagged code
4 flagged · loading sourcedist/discovery-B79IxEM6.cjsView file
1450package = @mulmoclaude/core; repositoryIdentity = mulmoclaude; dependency = @duckdb/node-api
L1450: async function duckDbInstance() {
L1451: if (instancePromise === null) instancePromise = import("@duckdb/node-api").then((mod) => mod.DuckDBInstance.create(":memory:"));
L1452: try {
High
Copied Package Dependency Bridge
Package metadata claims a different repository identity while copied source loads a runtime dependency bridge.
dist/discovery-B79IxEM6.cjsView on unpkg · L1450dist/skill-bridge/index.jsView file
•Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation.
import { t as issafeactiontemplatepath } from "../templatepath-k_wnbl_q.js";
import { mkdirsync, readfilesync, renamesync, rmsync, writefilesync } from "node:fs";
import path from "node:path";
var data_skills_dir = path.join("data", "skills");
var claude_skills_dir = path.join(".claude", "skills");
var sk[redacted] = "skill.md";
var schema_filename = "schema.json";
var slug_re = /^[a-z0-9]+(-[a-z0-9]+)*$/;
var rm_re = /^\s*rm\s+((?:-[a-za-z]+\s+)+)['"]?data\/skills\/([a-z0-9-]+)\/?['"]?\s*$/;
var recursive_flag_re = /[rr]/;
function dataskilldir(workspaceroot, slug) {
Findings
4 High3 Medium5 Low
HighCopied Package Dependency Bridgedist/discovery-B79IxEM6.cjs
HighTrigger Reachable External Ai Agent Control Surface Mutationdist/skill-bridge/index.js
HighSemantic Analysis Limiteddist/promptSafety-caJWd1ZQ.cjs#virtual:normalized:round1
HighPrevious Version Dangerous Deltadist/google/index.cjs
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowTelemetry
LowUrl Strings
Affected versions and remediation
This report applies to @mulmoclaude/core@5.4.0.
See all versions of @mulmoclaude/core for other recorded verdicts.
- Review the evidence and your use of @mulmoclaude/core@5.4.0 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.