@compr/opscontext-mcp@2.13.0
OpsContext for AI Agents — read-only fleet visibility (PM2/nginx/Docker/git/cron) + tamper-evident audit log + policy-as-code hooks. The ops + compliance layer Claude Code can't grow natively.
AI Security Review
scanned 12d ago · by lpm-firewall-aiThe source shows explicit user-command setup that can persist behavior in Claude configuration and the user's login session, so a capability warning is appropriate. No unconsented install-time mutation or other concrete attack behavior was established.
Static reason
No blocking static signals were detected.; previous stored version diff introduced dangerous source
Trigger
A user runs the corresponding install command.
Impact
These actions can make package functionality run during Claude sessions or user logins after explicit setup.
Mechanism
The Claude installer can add hooks to the user's Claude settings; the autostart installer writes and bootstraps a LaunchAgent plist. A separate installer copies a package-owned skill into a Claude skills directory.
AI rationale
The source shows explicit user-command setup that can persist behavior in Claude configuration and the user's login session, so a capability warning is appropriate. No unconsented install-time mutation or other concrete attack behavior was established.
Evidence
dist/cli.jsdist/install-claude-hook.jsdist/install-autostart.jsdist/claude-integration.js~/.claude/settings.jsonLaunchAgent plistClaude skill file
Decision evidence
public snapshotAI called this Suspicious at 84.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The manifest defines no preinstall, install, or postinstall hook; its prepublishOnly script runs during publishing.
- Explicit CLI commands invoke the Claude hook installer and autostart installer.
- The Claude hook installer can write hook configuration to the user's ~/.claude/settings.json.
- The autostart installer writes a LaunchAgent plist and bootstraps it for the user's login session.
- The bundled skill installer writes a package-owned skill file into a selected Claude skill directory.
Evidence against
- The inspected agent configuration changes are reached through explicit CLI commands, not an npm install lifecycle hook.
- The hook installer avoids writing settings when unchanged, and the skill installer protects an existing skill unless --force is supplied.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsFilesystemNetworkShell
Source & flagged code
7 flagged · loading sourcedist/adapters.jsView file
47}
L48: const mod = await import(moduleSpecifier);
L49: // Check for factory function
Medium
Dynamic Require
Package source references dynamic require/import behavior.
dist/adapters.jsView on unpkg · L47dist/install-autostart.jsView file
13Trigger-reachable persistence chain: manifest.bin -> dist/cli.js -> dist/install-autostart.js
L13: // forever — defeating the whole point.
L14: // FIX: To add platform support beyond macOS, branch on process.platform and
L15: // add equivalent systemd / NSSM logic. Keep `gui/$UID` and KeepAlive
...
L19: import { homedir, platform } from "os";
L20: import { execFileSync } from "child_process";
L21: import { createRequire } from "module";
...
L34: const __filename_esm = fileURLToPath(import.meta.url);
L35: const __dirname_esm = dirname(__filename_esm);
L36: const requireFromHere = createRequire(import.meta.url);
...
L102: // [LOCKED] [AUTOSTART-ARGV-AND-XML-ESCAPED] - 2026-09-25
L103: // [NEVER] paste a path into a launchctl/lsof/curl command string, or into the plist unescaped.
L104:
High
Trigger Reachable Persistence
A manifest entrypoint or package-local install chain reaches persistence behavior.
dist/install-autostart.jsView on unpkg · L13Findings
4 High6 Medium5 Low
HighTrigger Reachable External Ai Agent Control Surface Mutationdist/install-claude-hook.js
HighTrigger Reachable Persistencedist/install-autostart.js
HighSemantic Analysis Limiteddist/activation.js#virtual:normalized:round1
HighPrevious Version Dangerous Deltadist/cli.js
MediumDynamic Requiredist/adapters.js
MediumNetwork
MediumEnvironment Vars
MediumInstall Persistencedist/install-autostart.js
MediumShips Build Helperdefaults/claude-code-hook.sh
MediumStructural Risk Force Deep Review
LowNon Install Lifecycle Scripts
LowScripts Present
LowFilesystem
Affected versions and remediation
This report applies to @compr/opscontext-mcp@2.13.0.
See all versions of @compr/opscontext-mcp for other recorded verdicts.
- Review the evidence and your use of @compr/opscontext-mcp@2.13.0 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.