halfcycle@0.8.0
Your coding agents are the engineers. Halfcycle is the rest of the team: it scopes, designs, plans and reviews production software, and runs the build for Claude Code and Codex.
AI Security Review
scanned 21h ago · by lpm-firewall-aiThe reviewed network and credential flows support Halfcycle’s CLI and agent guard behavior, with service destinations or configuration controlling their use. The scanner’s reverse-shell lead corresponds to a loopback sign-in callback that checks the caller and state.
Static reason
High-risk behavior combination matched malicious policy.; previous stored version diff introduced dangerous source
Trigger
Users invoke CLI commands or configure the agent guard; importing the library exposes functions without invoking the CLI.
Impact
No unauthorized credential disclosure, unrelated execution, or other malicious behavior was found.
Mechanism
CLI requests use Halfcycle service credentials; the sign-in listener is restricted to loopback; CI OIDC and telemetry flows are tied to configured service endpoints.
AI rationale
The reviewed network and credential flows support Halfcycle’s CLI and agent guard behavior, with service destinations or configuration controlling their use. The scanner’s reverse-shell lead corresponds to a loopback sign-in callback that checks the caller and state.
Network endpoints2
control.halfcycle.aiguard.halfcycle.ai
Decision evidence
public snapshotAI called this Clean at 91.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The package exposes a library entry point and a CLI; its metadata declares no install lifecycle hook.
- The CLI’s sign-in callback accepts only loopback peers and checks a state value.
- The library’s engagement request sends its supplied bearer credential to the Halfcycle service URL.
- The guard’s CI identity flow requests a token from the job-provided OIDC URL and exchanges it with the configured control service.
- Guard telemetry is sent only when both a telemetry URL and token are configured.
Evidence against
- The CLI performs network requests and uses credentials when users invoke commands that need them.
- The package includes agent guard code that reads project context and can log guard evaluations.
Behavioral surface
ChildProcessCryptoEnvironmentVarsEvalFilesystemNetworkShell
HighEntropyStringsUrlStrings
Source & flagged code
11 flaggedbin/bin.bundle.mjsView file
•matchType = previous_version_dangerous_delta
matchedPackage = halfcycle@0.7.1
matchedIdentity = npm:aGFsZmN5Y2xl:0.7.1
similarity = 0.667
summary = stored previous version shares package body but lacks this dangerous source file
Critical
Previous Version Dangerous Delta
This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
bin/bin.bundle.mjsView on unpkg15163const program = [
L15164: "const p=require(`path`),f=require(`fs`),c=require(`child_process`);",
L15165: "let d=process.cwd();",
High
Child Process
Package source references child process execution.
bin/bin.bundle.mjsView on unpkg · L1516314667var WRITE_CAPABLE_TOOLS = ["Edit", "MultiEdit", "NotebookEdit", "Write"];
L14668: var PRE_TOOL_USE_MATCHER = "^(Bash|Read|Grep|Write|Edit|MultiEdit|NotebookEdit|Agent|PowerShell)$";
L14669: function hookCommand(script, verb) {
High
Findings
3 Critical6 High4 Medium6 Low
CriticalReverse Shelldist/cli.js
CriticalTrigger Reachable Dangerous Capabilitydist/cli.js
CriticalPrevious Version Dangerous Deltabin/bin.bundle.mjs
HighChild Processbin/bin.bundle.mjs
HighShellbin/bin.bundle.mjs
HighSame File Env Network Executiondist/cli.js
HighSandbox Evasion Gated Capabilitydist/cli.js
HighCross File Remote Execution Contextdist/cli.js
HighRuntime Package Installdist/cli.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperbin/pre-tool-use-filter.sh
MediumStructural Risk Force Deep Review
Affected versions and remediation
This report applies to halfcycle@0.8.0.
See all versions of halfcycle for other recorded verdicts.
- Review the evidence and your use of halfcycle@0.8.0 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.