The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute 'mandatory commands' and 'checks' that are specified within a human-approved implementation plan. This is a standard requirement for software development tasks and is governed by explicit instructions to seek human approval before execution.\n- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes external data from repository files. * Ingestion points: The agent reads from spec.md, plan.md, and project source files which may contain untrusted content. * Boundary markers: The instructions mandate verification of 'actual explicit human approval' and a comparison against an approved baseline to detect drift. * Capability inventory: The skill allows the agent to modify the filesystem and execute shell commands defined in the plan. * Sanitization: The skill relies on mandatory human-in-the-loop validation of the spec and plan to ensure no malicious instructions are processed or executed.

implement-change