The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several shell commands including find, ls, wc, git log, and cat to collect metadata and file contents from the project directory. All commands are targeted at a fixed absolute path: /Users/anthonybyrnes/PycharmProjects/Python419/.- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it reads and displays the body content of files that could potentially be influenced by external actors (e.g., TODOs and NEXT_STEPS files). * Ingestion points: Markdown files located in /docs/handoffs/, /docs/sessions/, and /docs/status/. * Boundary markers: The output format displays file content directly without specific delimiters or instructions to the agent to disregard embedded commands. * Capability inventory: The skill is limited to read-only operations and metadata extraction within the specified project structure. * Sanitization: The implementation uses yaml.safe_load for parsing frontmatter but does not apply sanitization to the main body of the markdown files.

project-status