The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The core functionality of the skill involves analyzing untrusted content from various files and Git history, which presents a surface for indirect prompt injection where malicious data could influence the agent's behavior.
Ingestion points: The skill instructs the agent to search through files such as appsettings.json, .env, Dockerfile, web.config, and GitHub workflow files.
Boundary markers: The instructions lack explicit delimiting markers to differentiate between the data being scanned and the agent's instructions.
Capability inventory: The skill utilizes the GitHub CLI (gh api) for alert management and Git tools (filter-repo, log) for history modification.
Sanitization: There are no specific instructions to sanitize or validate the content extracted from target files before the agent processes it.
[COMMAND_EXECUTION]: The skill provides functional templates for shell commands using gh api to interact with GitHub's secret scanning features and git filter-repo to perform history-rewriting remediation.

detecting-secrets