The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes rg (ripgrep) to search through agent session files located in ~/.pi/agent/sessions. This is a legitimate operation for the skill's purpose but involves shell command execution over internal data structures.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from historical session logs (.jsonl files) which may contain malicious instructions from previous user inputs or tool outputs.
Ingestion points: Reads files from ~/.pi/agent/sessions to reconstruct symptoms and trails.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified when processing the content.
Capability inventory: Uses rg for file discovery and diag-write for persistence.
Sanitization: The instructions do not specify any sanitization, escaping, or validation of the ingested session content before it is processed or archived.

diag-extract