agency-lang@0.7.0
The Agency language
AI Security Review
scanned 4mo ago · by lpm-firewall-aiStatic inspection found no lifecycle hook or import-time behavior that mutates broad AI-agent control surfaces, harvests credentials, or exfiltrates data. The flagged child_process, network, dynamic import, and persistence primitives are explicit CLI/runtime capabilities for user-authored Agency programs.
Decision evidence
public snapshot- package.json has no preinstall/install/postinstall lifecycle scripts; bin is user-invoked dist/scripts/agency.js.
- dist/lib/optimize/optimizerModule.js dynamically imports only a user-supplied optimizer file after CLI selection and removes the temp bundle.
- dist/lib/stdlib/oauth.js stores OAuth tokens under ~/.agency/oauth and posts only to user-configured HTTPS tokenUrl during explicit authorize/refresh flows.
- dist/lib/cli/schedule/backends/launchd.js creates launchd jobs only through agency schedule add/edit user commands.
- dist/lib/mcp/setup.js and dist/lib/lsp/setup.js write Codex/agent config only via explicit agency mcp/lsp setup commands, not install-time.
- dist/lib/runtime/providerModules.js imports provider modules from config/env paths as an advertised extensibility feature.
Source & flagged code
6 flaggedPackage source references child process execution.
dist/lib/optimize/grading/agencyRunner.jsView on unpkg · L31Findings
Affected versions and remediation
This report applies to agency-lang@0.7.0.
See all versions of agency-lang for other recorded verdicts.
- Review the evidence and your use of agency-lang@0.7.0 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.