The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local environment, including git history, existing source code, and PRD documents. This data is used to inform the generation of plans, designs, and analysis reports.
Ingestion points: Git history, project source code, and documentation files (e.g., docs/00-pm/{feature}.prd.md).
Boundary markers: The instructions do not specify explicit delimiters or "ignore instructions" warnings for the interpolated data.
Capability inventory: The skill uses Bash, Write, Edit, and Task tools, providing a significant impact surface if the agent obeys instructions embedded in the analyzed code.
Sanitization: No explicit sanitization or filtering of external content is described.
[DYNAMIC_EXECUTION]: During the analyze and qa phases, the skill generates executable test files (e.g., tests/e2e/{feature}-actions.spec.ts) and executes them using Playwright. It also constructs and executes curl commands for API verification based on an agent-generated plan.
[COMMAND_EXECUTION]: The skill leverages shell commands for project management tasks, such as environment detection (CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS), server status checks, and dependency installation (e.g., pnpm add -D @playwright/test).

pdca