The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell utilities such as grep, cut, head, and tail to parse and extract relevant information from local session transcripts and metadata files.
[COMMAND_EXECUTION]: It invokes the vendor's entire CLI tool to search session history, access current session data, and retrieve detailed checkpoint explanations.
[EXTERNAL_DOWNLOADS]: The entire search command is used to query the vendor's remote index for relevant workflow patterns based on user terms.
[PROMPT_INJECTION]: The skill processes potentially untrusted data from session transcripts and metadata files stored in the .git/entire-sessions/ directory.
Ingestion points: Content is ingested from local JSON session files and outputs from the entire explain command.
Boundary markers: No specific boundary markers or 'ignore' instructions are defined for the transcript processing steps.
Capability inventory: The skill can execute shell commands and write new SKILL.md files to the filesystem.
Sanitization: The instructions include a specific rule to prevent the inclusion of secrets, credentials, or raw logs in the final generated output.

session-to-skill