The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions in SKILL.md and references/quality-gates.md direct the agent to use the bash tool to execute curl commands as a fallback method for fetching web content when primary research tools are unavailable or restricted.
[DATA_EXFILTRATION]: The orchestration workflow involves reading local project documentation, architecture files, and existing research, then incorporating this potentially sensitive local context into mission briefs that are subsequently sent to internet-connected researcher agents.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it systematically fetches and processes large amounts of untrusted content from external sources (such as forums and community sites) and synthesizes this data into final documentation without explicit sanitization or boundary marking for the ingested content.

run-comprehensive-research