The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data in the form of code diffs.
Ingestion points: The skill executes git diff to gather changes across the worktree or specific branches for analysis (SKILL.md, Phase 0).
Boundary markers: There are no explicit instructions or delimiters provided to the cleanup agents to treat the content of the diff as inert data or to ignore embedded instructions.
Capability inventory: The skill uses git diff and grep for analysis, and it is explicitly authorized to apply fixes (file-write) based on the findings (SKILL.md, Phase 2).
Sanitization: The skill does not perform any sanitization, filtering, or validation of the diff content before it is processed by the cleanup agents.
[COMMAND_EXECUTION]: The skill executes local shell commands to manage its workflow and interface with version control.
Commands used: The skill runs git diff, grep, and echo (specifically echo simplify-guard:bypass and echo simplify-guard:complete).
Context: These commands are used for their intended purposes (diffing code, searching utilities, and signaling completion to a guard mechanism) and do not appear to incorporate unsanitized user input into the shell environment.

simplify