@markjaquith/agency@3.7.0
Manage agentic work across repositories with durable workbases
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThe automatic postinstall mutates a global AI-agent extension surface and the extension injects project instructions into agent prompts. The behavior is scoped to Agency's Pi extension and no concrete exfiltration or destructive behavior was found.
Decision evidence
public snapshot- The package automatically runs an extension installer after npm installation.
- The installer copies package code into the user's global Pi agent extension directory.
- The installed extension reads workbase instructions and appends them to Pi's system prompt.
- The manifest has no dependency on another release of itself.
- The reviewed lifecycle installer has no network or shell execution.
- No fixed exfiltration endpoint or credential harvesting was found in the inspected paths.
Source & flagged code
10 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgPackage source references dynamic require/import behavior.
src/utils/interactive-loader.tsView on unpkg · L3Findings
Affected versions and remediation
This report applies to @markjaquith/agency@3.7.0.
See all versions of @markjaquith/agency for other recorded verdicts.
- Review the evidence and your use of @markjaquith/agency@3.7.0 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.