The Agent Skills Directory

[SAFE]: The skill implements secure defaults by requiring HTTPS for webhook URLs and explicitly instructing the agent not to echo sensitive identifiers like Slack workspace IDs or full webhook URLs in its output.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through the ingestion and processing of external data, which is mitigated by specific sanitization instructions. 1. Ingestion points: The skill reads existing alert names and descriptions via the posthog:error-tracking-alerts-list tool (Step 2, SKILL.md). 2. Boundary markers: Uses standard curly-brace delimiters {} for property interpolation in templates. 3. Capability inventory: The skill can create or update alerts that send data to external integrations via posthog:error-tracking-alerts-create and posthog:error-tracking-alerts-partial-update (Step 4, SKILL.md). 4. Sanitization: Notification templates in references/block-templates.md mandate wrapping dynamic event properties (like issue descriptions) in markdown code blocks to prevent the rendering of malicious links, images, or mentions.

authoring-error-tracking-alerts