The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is primarily focused on researching and analyzing external data via WebSearch and WebFetch tools, creating an attack surface for indirect prompt injection.
Ingestion points: Untrusted content from pricing pages, changelogs, job postings, and community forums is ingested into the agent context for analysis.
Boundary markers: The skill refers to _common/WEB_FETCH_SAFETY.md for sanitization but does not define explicit delimiters in the local markdown files.
Capability inventory: Capabilities include subagent orchestration, strategic artifact generation, and network access for research.
Sanitization: SKILL.md mandates a prompt-injection check on every search result or fetch operation before the data is incorporated into reports.
[COMMAND_EXECUTION]: The skill references the use of standard developer tools such as git and gh (GitHub CLI) for intelligence gathering and profile auditing. These are legitimate capabilities within the scope of a technical research and branding tool.

compete