The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a process that ingests external, potentially untrusted data sources.
Ingestion points: In SKILL.md, the agent is instructed to "Establish the current problem from the work artifact, local Markdown, prototype, or code."
Boundary markers: The instructions do not mandate the use of delimiters or specific warnings to ignore instructions embedded within the ingested artifacts.
Capability inventory: The skill leverages the agent's ability to "build lightweight, reversible working prototypes" and "Writing and running code" to test behaviors, as noted in SKILL.md and references/development-principles.md.
Sanitization: There are no requirements for sanitizing or validating external content before it is interpolated into the agent's context or executed.
[DYNAMIC_EXECUTION]: The skill involves the generation and runtime execution of code for software exploration.
Evidence: SKILL.md directs the agent to "build... working prototypes in the actual product" and states that "Writing and running code can expose constraints the proposal missed." This dynamic creation and execution of scripts is the primary function of the skill.
[EXTERNAL_DOWNLOADS]: The skill references documentation from well-known technology organizations.
Evidence: references/writing-style.md contains links to official technical writing and voice guidelines hosted on official Google and Microsoft developer domains.

equilio-explore