The Agent Skills Directory

[DATA_EXFILTRATION]: The skill is designed to locate and read sensitive configuration and credential storage files, such as environment variables (.env), database schemas, and authentication implementations (JWT, passport), as part of its archaeological discovery process.
[COMMAND_EXECUTION]: The workflow involves using shell-based tools including Grep and Glob to perform recursive searches across the file system to map project structures and extract business logic.
[PROMPT_INJECTION]: There is a risk of indirect prompt injection because the skill ingest and interprets untrusted source code. Maliciously crafted comments or code content could attempt to override the agent's instructions or bias its documentation output.
Ingestion points: Project source code files accessed via Read, Grep, and Glob tools.
Boundary markers: No explicit markers or instructions are provided to the agent to distinguish between code being analyzed and operational instructions.
Capability inventory: Bash, Read, Grep, and Glob are utilized to process untrusted input.
Sanitization: The skill does not perform sanitization or validation of the content extracted from the analyzed codebase.

spec-miner