@kaydash9999/banker-plugins@0.19.2
banker — Claude Code + Codex CLI plugin (QA, audit, docs, architecture, wiki skills). Install via Claude Code marketplace, or `npm i -g @kaydash9999/banker-plugins` then `banker setup`.
AI Security Review
scanned 19d ago · by lpm-firewall-aiNo malicious install-time behavior was identified. The opt-in bypass-permissions capability presents a concrete security risk, supporting a warning rather than a publish block.
Static reason
No blocking static signals were detected.; previous stored version diff introduced dangerous source
Trigger
The user runs bypass-permissions on --yes.
Impact
Subsequent sessions can run tools without the usual confirmation prompts.
Mechanism
The helper writes bypassPermissions to the Claude Code settings file.
AI rationale
No malicious install-time behavior was identified. The opt-in bypass-permissions capability presents a concrete security risk, supporting a warning rather than a publish block.
Evidence
skills/setup-bypass-permissions/scripts/bypass-permissions.mjs~/.claude/settings.json
Decision evidence
public snapshotAI called this Suspicious at 88.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- package.json exposes a CLI and lists only a prepublishOnly version check; it has no npm install hook.
- The installer copies manifest-selected skills into user or project directories when the CLI setup command is run.
- The bypass-permissions skill can persistently set Claude Code's default mode to bypass tool confirmations, but accepts that change only with the explicit on --yes command.
- Other reviewed agent-related helpers are feature-specific: they modify local instruction or HUD files, while the capture helper validates user-provided targets and defaults to loopback.
Evidence against
- The bypass-permissions helper can remove tool-execution confirmations in later Claude Code sessions.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
6 flaggedbin/banker.jsView file
•Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation.
filesync(skillmd, src.slice(0, fm.index) + `---\n${block}\n---` + src.slice(fm.index + fm[0].length));
}
function setupclaude(dryrun) {
log('• claude code:');
if (!have('claude')) { warn(' claude cli not found on path — install claude code first. skipping --claude.'); return false; }
const steps = [
['claude', ['plugin', 'marketplace', 'add', pkg_root]],
['claude', ['plugin', 'install', `${plugin_name}@${marketplace}`]],
];
for (const [bin, args] of steps) {
if (dryrun) { log(` [dry-run] ${bin} ${args.join(' ')}`); continue; }
try {
High
Trigger Reachable External Ai Agent Control Surface Mutation
Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.
bin/banker.jsView on unpkg9*/
L10: const fs = require('fs');
L11: const os = require('os');
Medium
Dynamic Require
Package source references dynamic require/import behavior.
bin/banker.jsView on unpkg · L9skills/payload-mon/scripts/payload-size.mjs
Findings
3 High5 Medium6 Low
HighTrigger Reachable External Ai Agent Control Surface Mutationbin/banker.js
HighSemantic Analysis Limitedskills/payload-mon/scripts/payload-mon.mjs#virtual:normalized:round1
HighPrevious Version Dangerous Deltaskills/ultra-chain/references/capture.mjs
MediumDynamic Requirebin/banker.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperskills/audit-web-page/harness.py
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowWeak Cryptoskills/payload-mon/scripts/payload-size.mjs
LowFilesystem
LowHigh Entropy Strings
Affected versions and remediation
This report applies to @kaydash9999/banker-plugins@0.19.2.
See all versions of @kaydash9999/banker-plugins for other recorded verdicts.
- Review the evidence and your use of @kaydash9999/banker-plugins@0.19.2 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.