The Agent Skills Directory

[DATA_EXFILTRATION]: No network-capable operations or credential access patterns were identified. The skill's scope is strictly limited to local repository files.
[COMMAND_EXECUTION]: The skill does not utilize shell commands, subprocesses, or external scripts. It explicitly opts out of automated Git actions like add, commit, or push to ensure user control.
[PROMPT_INJECTION]: The provided instructions are purely functional and lack any patterns intended to bypass safety filters or override agent constraints.
[INDIRECT_PROMPT_INJECTION]: The skill processes repository files that may contain untrusted content.
Ingestion points: Git repository relative Markdown files used for saving and resuming discussions.
Boundary markers: The skill uses standard Markdown structure to parse data but does not employ specific security delimiters.
Capability inventory: Context is restricted to reading and writing Markdown files within the repository.
Sanitization: The risk is mitigated by the mandatory requirement to display proposed changes and wait for explicit user confirmation before any file creation or modification takes place.

discuss