The Agent Skills Directory

[PROMPT_INJECTION]: The skill directly interpolates user-provided $ARGUMENTS into the facilitation flow, instructing the agent that it "MUST consider the user input before proceeding". This creates a surface for direct prompt injection if the user provides malicious instructions intended to bypass governance rules.
[INDIRECT_PROMPT_INJECTION]: The skill ingests content from several team-controlled files to inform its principle proposals. If these files are compromised or contain malicious instructions, they could influence the agent's behavior.
Ingestion points: CDR.md, AGENTS.md, and files within {TEAM_AI_DIRECTIVES}/context_modules/rules/.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used when reading these files.
Capability inventory: The skill has the ability to write files (constitution.md) and execute git commit commands.
Sanitization: No explicit sanitization or filtering of the external file content is mentioned before it is processed by the model.
[COMMAND_EXECUTION]: The skill executes local setup scripts (scripts/bash/setup-team-constitution.sh and scripts/powershell/setup-team-constitution.ps1) and git commands for repository management. These are used for path resolution and state detection, which are standard operations for this type of tool.

team-constitution