The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/trace_session.py uses the subprocess module to execute git commands (e.g., git status, git log) to analyze repository state and extract session data.
[COMMAND_EXECUTION]: The scripts/install.sh script automates the setup of a post-commit git hook. This hook triggers the session tracer script after every commit to ensure log files remain synchronized with the repository state.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by reading untrusted data from git commit messages and file names, which are then incorporated into .cortex_log.md and .cortex_handoff.md files. If an agent subsequently reads these files as authoritative context, it could be influenced by malicious content embedded in the git history.
Ingestion points: scripts/trace_session.py reads git log and status output.
Boundary markers: Data is structured using Markdown headers and lists, but lacks explicit instruction-ignore delimiters.
Capability inventory: The skill utilizes file-system tools (Read, Write, Glob) and command execution via subprocess.
Sanitization: No specific sanitization or validation of the extracted git text is performed.

cortex