The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data such as git status and diff outputs which are interpolated into sub-agent prompts. While this represents a potential surface for indirect prompt injection if the source code contains malicious instructions, the risk is effectively mitigated by the skill's design.
Ingestion points: SKILL.md uses git status and git diff to collect session context which is passed to analysis agents.
Boundary markers: The skill does not use explicit delimiters to wrap interpolated content, but the agent roles are highly specialized.
Capability inventory: The skill can create commits, update CLAUDE.md, and generate new automation skills based on user selection.
Sanitization: The workflow includes a dedicated duplicate-checker validation agent and requires explicit user confirmation via AskUserQuestion before any modifications are performed.
[SAFE]: The skill uses platform-native orchestration tools (Task, AskUserQuestion) and standard version control commands (git status, git diff) for its primary functionality. External references point to well-known and trusted technology organizations (Anthropic, Microsoft) for educational purposes.

session-wrap