The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands for argument parsing, date calculation, and interacting with the GitHub CLI (gh). It includes input validation to ensure that the days argument is a positive integer before passing it to the date command.
[EXTERNAL_DOWNLOADS]: The skill fetches data from GitHub and Linear. These are well-known services, and the interactions are performed using authenticated tools (gh and Linear MCP) to retrieve the user's own data for the purpose of generating a summary.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, specifically PR titles and Linear ticket titles.
Ingestion points: Data enters the agent context via the output of gh search prs and Linear MCP tools.
Boundary markers: The final output is instructed to be wrapped in a 4-backtick outer fence to prevent inner markdown blocks from breaking the formatting.
Capability inventory: The skill uses Bash for data retrieval and Read/Write tools, though writing to files is only done upon explicit user request.
Sanitization: The skill applies transformations such as title-casing, emoji stripping, and jargon removal (in general audience mode) to the ingested data before rendering.

changelog