The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use the exec tool for informational shell commands, including listing directories (ls), searching for text (grep), and reading file contents (cat). These operations are restricted to the local proposals/ and decisions/ directories.
[PROMPT_INJECTION]: The skill includes operational constraints referred to as "Hard gates" and "STOP gates." These instructions are designed to prevent the agent from proceeding to solution design until the user has confirmed the problem framing, acting as a quality and safety guardrail for the agent's workflow.
[SAFE]: The skill presents a surface for indirect prompt injection because it reads and processes existing documentation files. This is consistent with its primary purpose of managing project-specific knowledge.
Ingestion points: proposals/ and decisions/ directories accessed via exec and search tools.
Boundary markers: None defined for the content of read files.
Capability inventory: exec, write, edit, and search tools.
Sanitization: No explicit content sanitization is described for read files.

open-knowledge-pack-software-lifecycle-frame-a-proposal