The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various git commands to manage development branches and worktrees.
Evidence:
git switch develop && git pull --ff-only && git switch -c feat/{slug}
git fetch origin develop && mkdir -p .worktrees && git worktree add .worktrees/{slug} -b feat/{slug} origin/develop
The skill also triggers external tools/commands like npm audit, pnpm audit, pip-audit, dotnet list package --vulnerable, and composer audit during the validation phase.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and follows instructions from untrusted external data sources, specifically project requirements and tickets.
Ingestion points: The skill reads docs/requirements/{slug}/TICKET.md, PRD.md, and other files in docs/ to determine functional requirements and implementation plans.
Boundary markers: The skill does not define explicit boundary markers or instructions to ignore embedded prompts within the ingested requirement files.
Capability inventory: The skill has the capability to write files, execute git commands, and trigger other automated tools based on the content of these documents.
Sanitization: No explicit sanitization or filtering of the content from the requirement documents is mentioned before it is processed by the agent.

fmway-dev