registry / @aspectlylabs/mikka / 0.0.439
@aspectlylabs/mikka@0.0.439
Local Mikka daemon: runs agents through the user's own CLIs
AI Security Review
scanned 6h ago · by lpm-firewall-aiThe package is a local daemon for running agents, with no install hooks in its manifest. The flagged credential and execution paths fit that purpose: model keys are routed through a broker to the configured provider, and GitHub access is scoped to run operations; no unrelated exfiltration or default account action was found.
Static reason
One or more suspicious static signals were detected.
Trigger
A user starts the daemon or a run using a configured model provider, or invokes GitHub operations through the run helper.
Impact
Requests and credentials reach the services selected for those features; no unrelated collection or automatic account action was found.
Mechanism
Model requests go through a local broker to the selected provider; GitHub access exchanges a run token with the configured server and limits the helper to GitHub HTTPS requests.
AI rationale
The package is a local daemon for running agents, with no install hooks in its manifest. The flagged credential and execution paths fit that purpose: model keys are routed through a broker to the configured provider, and GitHub access is scoped to run operations; no unrelated exfiltration or default account action was found.
Evidence
main.jsgithub-credentials.mjsmikka.js
Network endpoints4
api.anthropic.comapi.openai.comapp.aspectlylabs.comgithub.com
Decision evidence
public snapshotAI called this Clean at 92.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes a daemon CLI and module entry points; it lists no install lifecycle script.
- Provider credentials are validated for HTTPS and routed to the configured provider URL or the official Anthropic/OpenAI defaults.
- The API-key broker uses the selected provider base URL and rejects redirects when forwarding requests.
- The GitHub credential helper exchanges a run token with its configured endpoint using POST and rejects redirects.
- GitHub credentials are used by the helper only for GitHub HTTPS requests or an explicit gh mode.
Evidence against
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
CopyleftLicense
Source & flagged code
6 flaggedgithub-credentials.mjsView file
7import { randomUUID } from 'node:crypto'
L8: import { spawn } from 'node:child_process'
L9: import { pathToFileURL } from 'node:url'
High
Child Process
Package source references child process execution.
github-credentials.mjsView on unpkg · L7mikka.jsView file
4751ai: "7.0.142",
L4752: execa: "10.1.1",
L4753: pino: "10.4.0",
High
6760assertSafeId("run id", runId);
L6761: const common = resolve4(await worktreeGitDir(repoPath, worktree), "..", "..");
L6762: if (basename(common) !== ".git") throw new Error(`${repoPath} is not a standard clone`);
...
L6771: }
L6772: async function pinnedGit(repoPath, worktree, env = process.env) {
L6773: const gitDir = await worktreeGitDir(repoPath, worktree);
L6774: const workTree = realpathSync4(worktree);
L6775: return (args, extra, options) => execa3("git", [...SAFE_GIT, ...args], {
L6776: cwd: workTree,
Findings
5 High2 Medium5 Low
HighChild Processgithub-credentials.mjs
HighShellmikka.js
HighSame File Env Network Executionmikka.js
HighCredential Redirect Persistencemain.js
HighCommand Output Exfiltrationagents/native-container.mjs
MediumNetwork
MediumEnvironment Vars
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
LowCopyleft License
LowSuspicious Lifecycle Evidencepackage.json
Affected versions and remediation
This report applies to @aspectlylabs/mikka@0.0.439.
See all versions of @aspectlylabs/mikka for other recorded verdicts.
- Review the evidence and your use of @aspectlylabs/mikka@0.0.439 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.