The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads and processes multiple project files which may contain untrusted content that could influence the agent's behavior.
Ingestion points: The skill reads design/game-brief.md, GDD files in design/gdd/, architecture decisions in docs/architecture/adr-*.md, the control manifest, and the TR registry.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious instructions embedded within the design documents.
Capability inventory: The skill can create and update story files in the repository, execute local shell scripts via the Bash tool, and invoke the qa-lead agent for review.
Sanitization: No input sanitization or verification is performed on the data extracted from the project files.
[DYNAMIC_CONTEXT_INJECTION]: The skill uses the dynamic injection syntax (!) to execute a shell command at load time for configuration resolution. This command output is directly used to configure the skill's execution environment.
Evidence: !bash "${CLAUDE_SKILL_DIR}/../../hooks/yaml-helper.sh" resolve_config ...inSKILL.md`.
[COMMAND_EXECUTION]: The skill is configured with access to local scripts through the Bash tool to manage architecture status and document extraction. The tool definition allows for variable arguments to be passed to these scripts.
Evidence: Frontmatter allowed-tools includes definitions for yaml-helper.sh, adr-status.sh, and md-section.sh.

create-stories