The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill operates by reading and analyzing untrusted application source code from the project directory. This creates an attack surface where malicious instructions hidden in code comments or data could attempt to manipulate the agent's findings or execution flow.
Ingestion points: Files located within ${CLAUDE_SKILL_DIR}/, including source code, ORM models, and database queries.
Boundary markers: The instructions do not specify any delimiters or safety prompts to isolate external content from the agent's internal logic.
Capability inventory: The skill is granted Write, Edit, and Bash(security-test:*) permissions.
Sanitization: No sanitization or validation of the analyzed file content is documented.
[COMMAND_EXECUTION]: The skill is configured to use the Bash(security-test:*) tool. It references bundled scripts such as sqli_scan.py and sqli_exploit.py for automated vulnerability detection and verification, which involves executing code to simulate database attacks.
[SAFE]: The skill includes links to intentsolutions.io, tonsofskills.com, and jeremylongshore.com. These are identified as legitimate resources associated with the skill's author.

detecting-sql-injection-vulnerabilities