The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various shell commands to manage files and run validation tools, including mkdir, rsync, and cp for template management. It also invokes local Python scripts within the repository, such as .github/actions/gate/run.py and scripts/build_docs.py, to verify service definitions and documentation status.\n- [DYNAMIC_EXECUTION]: The skill uses an inline Python script passed via a heredoc to parse local YAML alert rules. This script audits the rules for severity levels and the presence of runbook links without executing any code from the files themselves.\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests data from external sources to generate its output. \n
Ingestion points: reads from services/*.yml, Prometheus alert rule files, and an optional directory specified by the $POWERSUIT_OVERLAY environment variable. \n
Boundary markers: None present to delimit untrusted data from instructions. \n
Capability inventory: Includes file system writes (writing docs/sre/on-call.md), shell command execution, and running local Python scripts. \n
Sanitization: No sanitization is performed on the ingested content before it is interpolated into the generated on-call policy document.

on-call