The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external bug reports (Jira and GitHub) to drive its diagnostic and repair workflows.
Ingestion points: External data is ingested from Jira fields (summary, description, components) and GitHub issues (body, labels) as described in workflows/fix-bug.md (Step 1).
Boundary markers: There are no explicit instructions to use delimiters or ignore instructions embedded within the ingested issue data when generating code or tests.
Capability inventory: The skill possesses significant capabilities, including the ability to write files (reproduction tests and product code) and execute shell commands via npx playwright, git, and the GitHub CLI (gh).
Sanitization: The workflow does not specify sanitization or escaping mechanisms for the external data before it is interpolated into reproduction tests (_repro-<issue>.test.ts) or diagnostic reports.
[COMMAND_EXECUTION]: The skill relies on various shell commands to perform its duties, including git remote, yarn start, npx playwright test, and gh api. It also includes logic to discover and potentially terminate processes owning specific network ports in workflows/fix-bug.md (Step 4A).
[EXTERNAL_DOWNLOADS]: The workflow allows for the execution of the repository's documented installation commands (e.g., yarn install) if dependencies are found to be missing or stale (Step 3 in workflows/fix-bug.md). It also references the installation of system utilities like ffmpeg via brew or apt.
[DYNAMIC_EXECUTION]: The skill dynamically generates Playwright test files (_repro-*.test.ts) based on issue descriptions and then executes them to verify bugs and fixes. This is a core functional requirement of the skill but involves executing newly created code.

rhdh-plugin-bug-fix