The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and analysis of external Go source code, which could serve as a vector for indirect prompt injection.
Ingestion points: The skill instructs the agent to search and read project files (*.go) using Glob, Grep, and Read tools.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' directives when processing the content of audited files, which could lead the agent to follow instructions embedded in code comments.
Capability inventory: The agent is granted Write and Edit permissions, as well as Bash access for Go toolchain commands (e.g., go test, go build), which could be misused if an injection occurs.
Sanitization: The skill lacks procedures for sanitizing or escaping the content of analyzed files before it is processed by the agent.

go-concurrency-review