asc-plugin-release
Fail
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: HIGHINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a
manifest.jsonfile (including name, version, and description) and interpolates these values directly into shell commands for GitHub releases (gh release create). - Ingestion points:
manifest.json(Step 1). - Boundary markers: None. Values are directly used in command strings.
- Capability inventory:
make build,zip,gh release,git commit,git push(Steps 2-7). - Sanitization: None specified. If a plugin manifest contains shell metacharacters in its metadata fields, it could potentially trigger unintended command execution during the release process.
- [COMMAND_EXECUTION]: The skill relies on executing local shell commands (
make build,zip,gh,git) to build and publish plugins. While appropriate for the skill's stated purpose, these represent a high-privilege capability tier that requires monitoring. - [REMOTE_CODE_EXECUTION]: Analysis confirms that the automated scanner alert regarding remote code execution is a false positive. The command
curl -s ... | python3 -m json.toolin Step 8 uses the Pythonjson.toolmodule specifically to format and validate JSON output, not to execute the downloaded content as code. Furthermore, the source URL belongs to the skill vendor's own infrastructure.
Recommendations
- HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/tddworks/asc-registry/main/registry.json - DO NOT USE without thorough review
Audit Metadata