tech-lead-setups
Warn
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions in
references/anti-slop.mdto download source code from an external repository (github.com/dmmulroy/anti-slop) and copy it into the project's own directory tree. This source is not associated with a trusted organization or well-known service, and manual source-code copies bypass standard dependency auditing tools. - [COMMAND_EXECUTION]: The skill performs extensive shell operations including project initialization, dependency installation, and a verification suite that runs user-defined check commands (
<type-check> && <lint> && <format-check> && <test> && <build>). These commands are executed with parameters and names derived from configuration files and documentation found within the repository. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from untrusted sources in the repository to determine project structure, architectural patterns, and execution parameters.
- Ingestion points: Documentation files including
docs/technical-specs/,docs/task-breakdown/,docs/api-specs/, andCLAUDE.md. - Boundary markers: The instructions do not define explicit boundary markers or use "ignore instructions" warnings when processing these files.
- Capability inventory: The skill possesses the ability to write arbitrary files to the disk and execute shell commands for project initialization and verification.
- Sanitization: There is no evidence of sanitization or validation for the directory names, pattern names, or tool configurations derived from the documentation files before they are used in file system or shell operations.
Audit Metadata