The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided Markdown files within the .product/spec/ directory to synthesize a consolidated specification. It lacks explicit instructions to treat this content as purely data, which creates a potential surface for indirect prompt injection if malicious instructions are embedded in the source files.\n
Ingestion points: The skill reads V*.md and CURRENT.md files from the local repository during sync, check, and export operations.\n
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded commands within the ingested files.\n
Capability inventory: The skill utilizes Bash, Write, Edit, and Read tools, providing a significant capability set if an injection were successful.\n
Sanitization: While the skill validates file metadata and naming, it does not describe sanitization or filtering of the Markdown content itself.\n- [COMMAND_EXECUTION]: The skill relies on the Bash tool to perform repository reconnaissance and file operations. The usage of user-provided strings (via $ARGUMENTS) in shell contexts introduces a potential command injection risk, although the skill instructions include logic for generating safe slugs from titles.

product-spec-generate