The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to perform architectural reconnaissance, searching for infrastructure configurations and application entry points using tools like find and grep.
[CREDENTIALS_UNSAFE]: The instructions direct the agent to identify sensitive authentication markers such as jwt, oauth, and token within the codebase. This access to credentials is intended for security analysis and threat identification as defined in the skill's purpose.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests and processes untrusted data from local codebase files. Malicious content within scanned files could theoretically influence the agent's analysis.
Ingestion points: Files read and analyzed during the 'Step 0' codebase scan (SKILL.md).
Boundary markers: The instructions do not define specific delimiters for separating codebase content from agent instructions.
Capability inventory: The skill has access to Bash, WebFetch, WebSearch, and file manipulation tools (Write, Edit).
Sanitization: There is no explicit logic to sanitize or escape data retrieved from the codebase before it is processed by the agent.

warden-threat