The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, creating an attack surface for indirect instructions.
Ingestion points: The agent reads CI failure logs (gh run view --log-failed) and Allure test result JSON files, which may contain attacker-influenced content from the application under test.
Boundary markers: There are no explicit instructions for using delimiters or boundary markers when parsing log content to differentiate data from instructions.
Capability inventory: The skill possesses significant capabilities, including triggering GitHub workflows (gh workflow run), setting repository secrets (gh secret set), and interacting with Jira via the acli tool.
Sanitization: The instructions do not detail specific sanitization or filtering logic for external log data before it is processed by the agent.
[COMMAND_EXECUTION]: The skill relies heavily on shell command execution to perform its primary tasks.
System Interaction: Uses the GitHub CLI (gh) to monitor runs, download artifacts, and manage repository state.
Local Scripts: Invokes project-specific scripts such as bun allure:agent, bun run test:sync, and scripts/ci/publish-allure-pages.ts.
Administrative Tasks: Guides the user and agent through administrative CLI tasks for configuring GitHub Pages, Vercel deployments, and Supabase projects.
[EXTERNAL_DOWNLOADS]: The skill fetches resources and uses tools from well-known and trusted external services.
Artifact Retrieval: Downloads Allure results and failure evidence (screenshots/traces) from GitHub Actions runners.
Tool Dependencies: References and uses established CLI tools including allure, wrangler (Cloudflare), vercel, and supabase for environment setup and reporting.
Official Registries: Recommends package updates for allure-playwright and allure-js-commons from official registries.

regression-testing