The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard command-line utilities to gather project context and manage internal logs.
Evidence: Executes date, git rev-parse, find, ls, and mkdir to identify the project root and organize prediction files within the Έ/ directory.
[EXTERNAL_DOWNLOADS]: The skill references common developer tools for repository scanning without performing automated installations.
Evidence: Mentions the use of ghq for cross-repository pattern discovery and a local script dig.py for session history analysis.
[PROMPT_INJECTION]: The skill analyzes external data such as git logs and session history, which constitutes a potential surface for indirect prompt injection.
Ingestion points: Reads from git log, ghq managed repositories, and session history files.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are defined for the ingested data.
Capability inventory: Includes the ability to execute shell commands for analysis and write findings to the project's metadata directory.
Sanitization: No explicit sanitization or filtering of the ingested content is described in the logic.

morpheus