The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to verify the local Python virtual environment by attempting to import the 'mcp' module and checking for the existence of configuration files.
[DYNAMIC_EXECUTION]: In Step 4.5, the skill is instructed to process plugin upgrades by executing 'auto' actions (shell commands) provided by the 'get_pending_migrations' MCP tool. While these are intended as vendor-provided migration scripts, they constitute the execution of dynamically generated commands.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project-specific files which could lead to unauthorized behavior.
Ingestion points: The skill reads instructions from {overrides}/CLAUDE.md and {overrides}/pronunciation-guide.md during Step 3.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present when reading these files.
Capability inventory: The agent has access to Bash and bitwize-music-mcp tools, which can be used to modify the local filesystem or execute commands.
Sanitization: There is no evidence of sanitization or validation of the content read from the override files before it is incorporated into the agent's instructions.

session-start