The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses WebSearch and WebFetch to retrieve information from external websites. This is the core functionality for researching technical options, benchmarks, and community health.
[PROMPT_INJECTION]: The instructions use emphatic, repetitive language (e.g., 'MANDATORY IMPORTANT MUST ATTENTION') to enforce a strict workflow. This is intended to ensure task breakdown, user interaction, and evidence-based reporting, rather than attempting to bypass safety filters or extract system prompts.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from web searches and local artifacts, creating a potential surface for indirect injection if that data contains malicious instructions.
Ingestion points: External content from WebFetch and local business context files in plans/ and team-artifacts/.
Boundary markers: Not explicitly defined for external content.
Capability inventory: The skill has access to Bash, Write, and Edit tools.
Sanitization: No explicit sanitization or instruction-filtering logic is present for the ingested data.

tech-stack-research