tracist@0.0.60
Upload and continuously stream local AI coding transcripts to Tracist
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is an overt, authenticated product feature rather than stealth install-time malware, but it grants a remote service unusually broad control over local AI agents. Warn rather than block because persistence is explicit-user-command and no unconsented lifecycle mutation or covert exfiltration was found.
Decision evidence
public snapshot- scripts/capture-resume-multiturn.js accepts authenticated queue messages and dispatches prompts to local agent sessions.
- It starts/resumes Codex with sandbox "danger-full-access" and approvalPolicy "never".
- The explicit install command creates a persistent user daemon that maintains the remote WebSocket control channel.
- Queued attachments are downloaded from the service and written locally before being included in agent prompts.
- package.json has only a prepare build hook; no preinstall/install/postinstall runtime hook.
- bin/tracist.js requires an explicit "tracist install" command before creating the daemon.
- Transcript uploads are declared functionality and outbound transcript rows pass through tools/secret-redaction.js.
- No hidden endpoint, credential harvesting, or foreign AI-agent configuration mutation was found.
Source & flagged code
8 flagged · loading sourceA manifest entrypoint or package-local install chain reaches persistence behavior.
bin/tracist.jsView on unpkg · L7Findings
Affected versions and remediation
This report applies to tracist@0.0.60.
See all versions of tracist for other recorded verdicts.
- Review the evidence and your use of tracist@0.0.60 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.