The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions for the user to download and execute an installation script for the traceway-write CLI from https://cli.tracewayapp.com/install.sh. This is an official setup method provided by the vendor.
[COMMAND_EXECUTION]: The skill relies on executing shell commands through the traceway-write CLI or using MCP tools to modify resources such as archiving issues, resolving pages, and snoozing alerts. It includes safety practices like human-in-the-loop confirmations before making changes.
[INDIRECT_PROMPT_INJECTION]: The skill processes dashboard URLs provided by users to perform lookups. Parsing data from external, user-supplied URLs represents a potential attack surface where instructions could be embedded.
Ingestion points: User-pasted dashboard URLs (e.g., https://<instance>/issues/<hash>) handled in SKILL.md workflows.
Boundary markers: The instructions provide specific mapping rules for various URL patterns but do not explicitly detail isolation techniques for the URL content itself.
Capability inventory: The skill can execute various write operations (archive, resolve, snooze, create post-mortems) via the CLI and MCP tools.
Sanitization: Instructions mention URL-decoding segments and require the presence of specific timestamp parameters for efficiency.

traceway-write