The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The instructions direct the agent to run npx claude-code-templates@latest --agent {category}/{name} --yes. This pattern downloads and executes code from the public NPM registry at runtime. The use of the --yes flag automatically accepts the package installation, removing a manual confirmation step for the user.
[COMMAND_EXECUTION]: The skill relies on a suite of commands using the bd CLI to manage a task lifecycle (epic, task, dependency, claim, close). It also requires the use of git for worktree management and branch finalization.
[DYNAMIC_EXECUTION]: The skill utilizes npx to dynamically fetch and execute specialist agent templates at runtime. The specific code executed is determined by variables provided during the workflow.
[INDIRECT_PROMPT_INJECTION]: The skill defines a process that ingests external, potentially untrusted data.
Ingestion points: External task definitions passed to bd create and library documentation retrieved via mcp__context7__query-docs in the SKILL.md file.
Boundary markers: Absent. No delimiters are specified to isolate external task text or documentation content from the agent's internal instructions.
Capability inventory: The workflow involves shell access (bd, git, npx), file writes during TDD implementation, and execution of planning and debugging logic.
Sanitization: Absent. The skill does not describe any validation or filtering of external inputs before they are used to drive the brainstorming and planning phases.

unified-workflow