The Agent Skills Directory

[COMMAND_EXECUTION]: The skill (in SKILL.md) executes shell commands, including git remote -v and git check-ignore, to identify hosting providers and manage repository ignore rules.\n- [INDIRECT_PROMPT_INJECTION]: The skill reads various repository files to inform its configuration process, which creates a potential surface for indirect prompt injection if those files contain adversarial instructions.\n
Ingestion points: SKILL.md instructs the agent to read CLAUDE.md, AGENTS.md, CONTEXT.md, CONTEXT-MAP.md, and Architecture Decision Records (ADRs) within docs/adr/.\n
Boundary markers: Absent; no explicit delimiters or instructions to ignore embedded content are specified for the read operations.\n
Capability inventory: SKILL.md performs file writes to create or update docs/agents/ documentation, CLAUDE.md, AGENTS.md, and the root .gitignore file, and executes Git commands.\n
Sanitization: Present; the skill is designed to present findings and draft content to the user for confirmation before performing any write operations.

tracer-setup