The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references and installs the ecc-agentshield package from the NPM registry and documentation from GitHub (affaan-m/agentshield). These are central to the skill's functionality and align with the ECC origin specified in the metadata.
[REMOTE_CODE_EXECUTION]: The skill uses npx ecc-agentshield to execute the scanner at runtime. This involves fetching and running code from the NPM registry, which is a standard execution method for this CLI-based security tool.
[COMMAND_EXECUTION]: The skill requires shell access to execute NPM commands, installation scripts, and the scanning engine (ecc-agentshield scan).
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and process untrusted project data including CLAUDE.md, .claude/settings.json, and agent definitions.
Ingestion points: Files located in the .claude/ directory and project root.
Boundary markers: Not explicitly defined in the skill instructions.
Capability inventory: Subprocess execution via shell commands (npx, npm).
Sanitization: The skill itself functions as a security scanner intended to detect and mitigate prompt injection and other risks within the processed files.

security-scan