registry / autopilot-ai / 2.8.0
autopilot-ai@2.8.0
autopilot (formerly kimiflare): a terminal coding agent for any model with your own key: OpenRouter, Anthropic, OpenAI, Gemini, Bedrock, Vertex, Ollama and more. Per-turn cost, web search, GitHub, headless browser, LSP, MCP, local memory.
AI Security Review
scanned 7d ago · by lpm-firewall-aiThe reviewed signals map to normal AI-provider routing and an isolated code runner for a terminal coding agent. The manifest shows no install hook that contacts an unrelated host, and the source does not establish a concealed credential redirect or exfiltration path.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
The user starts the CLI, selects a provider, or requests an agent task.
Impact
No unauthorized credential forwarding, unrelated command-output exfiltration, or install-time contact was identified.
Mechanism
Provider credentials are paired with the configured provider; agent code runs in an isolated VM with the supplied tool bridge.
AI rationale
The reviewed signals map to normal AI-provider routing and an isolated code runner for a terminal coding agent. The manifest shows no install hook that contacts an unrelated host, and the source does not establish a concealed credential redirect or exfiltration path.
Network endpoints1
ai-gateway.vercel.sh/v1
Decision evidence
public snapshotAI called this Clean at 94.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest exposes a terminal CLI and library entrypoints; its only lifecycle script is prepublishOnly, which runs the build.
- Vercel AI Gateway is a named provider with its own required AI_GATEWAY_API_KEY and Vercel dashboard URL.
- Provider resolution selects a provider from explicit environment or saved configuration and uses that provider’s key and endpoint.
- The isolated VM executes supplied code with a tool bridge built from the tools explicitly passed to it.
Evidence against
- The bundle includes process execution and network access, but these fit the stated purpose of a terminal coding agent.
- The scanner’s gateway and dynamic-execution signals correspond to provider configuration and the agent’s isolated code runner; no unrelated credential forwarding was identified.
Behavioral surface
ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNetworkShell
Source & flagged code
15 flaggeddist/index.jsView file
41Source sends the broad process environment to a literal external destination.
L41: function defaultLogDir() {
L42: const xdg = process.env.XDG_CONFIG_HOME || join(homedir(), ".config");
L43: return join(xdg, "kimiflare", "logs");
...
L58: if (currentStream) {
L59: currentStream.end();
L60: currentStream = null;
...
L188: function setFetchForTesting(f) {
L189: fetchImpl = f ?? ((url, init) => fetch(url, init));
L190: }
...
L989: import { createSign } from "crypto";
L990: import { execFile } from "child_process";
L991: import { readFile } from "fs/promises";
Critical
Hardcoded Runtime Data Exfiltration
Source sends credentials or rich application records to a package-controlled external receiver enabled by default.
dist/index.jsView on unpkg · L411055const iat = Math.floor(now2 / 1e3);
L1056: const payload = enc({ iss: sa.client_email, scope: SCOPE, aud: sa.token_uri ?? "https://oauth2.googleapis.com/token", iat, exp: iat + 3600 });
L1057: const signature = createSign("RSA-SHA256").update(`${header}.${payload}`).sign(sa.private_key).toString("base64url");
...
L1062: const tokenUrl = creds.type === "service_account" ? creds.token_uri ?? "https://oauth2.googleapis.com/token" : "https://oauth2.googleapis.com/token";
L1063: const body = creds.type === "service_account" ? new URLSearchParams({ grant_type: "urn:ietf:params:oauth:grant-type:jwt-bearer", assertion: signJwt(creds, now2) }) : new URLSearchP...
L1064: const res = await fetchImpl2(tokenUrl, { method: "POST", headers: { "Content-Type": "application/x-www-form-urlencoded" },
Critical
Command Output Exfiltration
Findings
6 Critical8 High4 Medium6 Low
CriticalHardcoded Runtime Data Exfiltrationdist/index.js
CriticalCommand Output Exfiltrationdist/index.js
CriticalDns Exfiltrationdist/index.js
CriticalRemote Response Code Executiondist/index.js
CriticalTrigger Reachable Dangerous Capabilitydist/index.js
CriticalPrevious Version Dangerous Deltadist/sdk/index.d.ts
HighChild Processdist/index.js
HighShelldist/index.js
HighEval
HighSame File Env Network Executiondist/index.js
HighCredential Redirect Persistencedist/index.js
HighRemote Agent Bridgedist/index.js
Affected versions and remediation
This report applies to autopilot-ai@2.8.0.
See all versions of autopilot-ai for other recorded verdicts.
- Review the evidence and your use of autopilot-ai@2.8.0 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.