The Agent Skills Directory

[COMMAND_EXECUTION]: The skill defines bash scripts that utilize curl to interact with the Rootly API endpoints. These scripts use standard Unix utilities such as jq, column, sort, and uniq to filter, format, and display incident and service information from JSON responses.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by ingesting and displaying data from an external API.
Ingestion points: Data from the Rootly API (e.g., incident titles, summaries, and timeline events) is retrieved and output to the agent's context in SKILL.md during the discovery and analysis phases.
Boundary markers: The skill does not use specific delimiters or warnings to isolate external data from the agent's primary instructions.
Capability inventory: The skill is restricted to making network calls to the official Rootly API (api.rootly.com) and processing the results. It does not have access to sensitive local files or arbitrary command execution capabilities.
Sanitization: The skill employs jq to truncate certain fields (e.g., .attributes.title[0:60] and .summary[0:200]), which provides basic length-based filtering of external content.

managing-rootly