@darkhunt-security/endpoint-claude-code@0.9.6
Darkhunt endpoint adapter for Claude Code: hook codec, transcript mapper, plugin manifest.
AI Security Review
scanned 4d ago · by lpm-firewall-aiThe package describes its purpose as Claude Code trace capture and guardrails, and the plugin manifest discloses that capture is on by default. The reviewed credential flow targets the configured Darkhunt API, while the forwarder enables masking by default; the scanner flags do not establish malicious behavior.
Static reason
One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source
Trigger
Trace capture occurs when the adapter is configured and used; enrollment runs through its explicit CLI command.
Impact
Expected impact is transmission of session trace data to the configured Darkhunt service; no unrelated credential collection or concealed execution was found.
Mechanism
The forwarder posts authenticated telemetry to the configured Darkhunt trace endpoint, with masking enabled by default.
AI rationale
The package describes its purpose as Claude Code trace capture and guardrails, and the plugin manifest discloses that capture is on by default. The reviewed credential flow targets the configured Darkhunt API, while the forwarder enables masking by default; the scanner flags do not establish malicious behavior.
Network endpoints2
api.darkhunt.aiDARKHUNT_BASE_URL (configurable trace endpoint)
Decision evidence
public snapshotAI called this Clean at 93.0% confidence as Benign with low false-positive risk.
Evidence for AI clean decision
- The manifest describes the package as a Darkhunt endpoint adapter for Claude Code.
- The plugin manifest discloses that full session trace capture is on by default and enforcement is off by default.
- The enroll CLI accepts a supplied key, reads DARKHUNT_API_KEY, or reads Darkhunt CLI credentials.
- Enrollment sends the key to the configured API base, defaulting to https://api.darkhunt.ai.
- The forwarder enables capture by default and sends authenticated requests to its configured trace endpoint.
- The telemetry sanitizer redacts values according to its ruleset, and masking is enabled by default.
Evidence against
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsUrlStrings
NoLicense
Source & flagged code
36 flaggeddist/bin/forwarder.mjsView file
19204patternName = github_pat
severity = critical
line = 19204
Critical
Critical Secret
Package contains a critical-looking secret pattern.
dist/bin/forwarder.mjsView on unpkg · L1920419204patternName = github_pat
severity = critical
line = 19204
Critical
Secret Pattern
GitHub personal access token in dist/bin/forwarder.mjs
dist/bin/forwarder.mjsView on unpkg · L1920419205patternName = github_oauth
severity = critical
line = 19205
Critical
Secret Pattern
GitHub OAuth access token in dist/bin/forwarder.mjs
dist/bin/forwarder.mjsView on unpkg · L1920519224patternName = slack_bot_token
severity = critical
line = 19224
Findings
20 Critical11 High8 Medium4 Low
CriticalCritical Secretdist/bin/forwarder.mjs
CriticalPrevious Version Dangerous Deltadist/bin/status.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/forwarder.mjs
CriticalSecret Patterndist/bin/backfill.mjs
Affected versions and remediation
This report applies to @darkhunt-security/endpoint-claude-code@0.9.6.
See all versions of @darkhunt-security/endpoint-claude-code for other recorded verdicts.
- Review the evidence and your use of @darkhunt-security/endpoint-claude-code@0.9.6 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.