The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell scripts using bash scripts/fetch-cc-changelog.sh as defined in SKILL.md and utilizes grep -r for recursive searches in the file system as described in references/analysis-rules.md.
[EXTERNAL_DOWNLOADS]: The skill is designed to fetch external data (the Claude Code changelog) via a script, although the specific URL is not explicitly defined in the provided instruction files.
[PROMPT_INJECTION]: The skill processes untrusted external data which creates a surface for indirect prompt injection. 1. Ingestion points: External content is ingested from the output of scripts/fetch-cc-changelog.sh mentioned in SKILL.md. 2. Boundary markers: Absent; SKILL.md does not specify delimiters to isolate fetched content. 3. Capability inventory: Shell script execution in SKILL.md and file system searching in references/analysis-rules.md. 4. Sanitization: Absent; the skill does not implement validation or filtering of the fetched changelog entries before they are analyzed by the agent.

cc-changelog