The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard command-line tools such as git, gh (GitHub CLI), and date to gather project history and metadata. These operations are essential for the skill's purpose and do not include arbitrary or high-risk command execution patterns.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external sources, including pull request titles, commit messages, issue descriptions, and changelog files. While this creates a theoretical surface for indirect prompt injection, the risk is minimal because the output is a plain-text draft intended for human review and the skill contains explicit instructions to remain in a research/drafting state.
Ingestion points: Commit messages via git log, pull request data via gh pr list, issue data via gh issue list, and local files such as CHANGELOG.md and roadmap.md (specified in SKILL.md and references/setup.md).
Boundary markers: The skill does not define specific delimiters or isolation markers to wrap untrusted content when presenting it to the agent.
Capability inventory: The skill is capable of executing shell commands (Git/GitHub CLI) and reading/writing local configuration and log files.
Sanitization: No explicit sanitization or filtering of the ingested external data is performed before processing.

weekly-progress-note