The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from stakeholder interviews and script parameters to generate Confluence templates, creating a vector where poisoned input could influence published pages.\n
Ingestion points: Discovery workflows in SKILL.md and the --sections or --macros parameters in scripts/template_scaffolder.py.\n
Boundary markers: The skill explicitly instructs the agent to "confirm template renders correctly in preview before publishing" and "verify each create before proceeding."\n
Capability inventory: Uses mcp__atlassian__createConfluencePage and mcp__atlassian__updateConfluencePage to commit generated markup to the target environment.\n
Sanitization: The Python scaffolder script uses direct string interpolation for markup generation without explicit escaping of input values.

atlassian-templates