The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from external sources (Jira API issues and local Markdown documents) to generate traceability reports. It lacks explicit boundary markers or sanitization to prevent indirect prompt injection if the source data contains embedded instructions aimed at the agent.
Ingestion points: The extract_jira_requirements function in SKILL.md fetches issue summaries and descriptions from Jira; extract_markdown_requirements in SKILL.md parses local Markdown files.
Boundary markers: No delimiters or "ignore embedded instructions" markers are used when processing the text content from external sources.
Capability inventory: The skill uses requests.get in SKILL.md for network access and os.walk/open in SKILL.md for file system access to build and export reports.
Sanitization: The implementation does not perform validation or filtering on the content extracted from requirements sources before using it in the application logic.
[COMMAND_EXECUTION]: The instructions include a shell script for a git pre-commit hook and a YAML configuration for GitHub Actions. These are standard development tools used to automate traceability checks and enforce documentation standards within a repository.
[EXTERNAL_DOWNLOADS]: The skill references the requests library for Python and official GitHub Actions (actions/checkout, actions/upload-artifact). These are well-known resources being used for their intended purpose of fetching API data and managing build artifacts.

traceability-matrix-generator