The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes bash and jq for directory management and updating the project manifest. These commands are applied to local project files and utilize secure argument-passing techniques and variable quoting to mitigate injection risks.
[PROMPT_INJECTION]: An indirect prompt injection surface exists because the skill reads and processes user-provided markdown files to generate agent instructions. Ingestion points: Reads all files in the docs/prds/ directory. Boundary markers: Absent; the skill does not use specific delimiters or warnings to separate PRD content from rule templates. Capability inventory: The skill has Write access to the .claude/rules/ directory and uses Bash to modify docs/blueprint/manifest.json. Sanitization: Absent; extracted content is reformatted directly into rules.
[SAFE]: No malicious patterns, obfuscation, or unauthorized access to sensitive system paths or credentials were identified. All activities are confined to the local project structure.

blueprint-generate-rules