The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests test results from playwright-results.json during the validation phase. These results may contain untrusted data from the application under test, such as DOM snapshots or error logs. The 'LLM triage' step processes this content without explicit boundary markers or sanitization, creating a surface where malicious application content could influence agent behavior.\n
Ingestion points: playwright-results.json (SKILL.md, Phase 4).\n
Boundary markers: Absent.\n
Capability inventory: Bash, Write, Edit, Task, Skill.\n
Sanitization: Absent.\n- [COMMAND_EXECUTION]: The instructions require the agent to execute several shell commands to set up the environment, verify compilation, and run tests.\n
Evidence: npx playwright test, npx playwright install, and npx tsc --noEmit in SKILL.md.\n- [DYNAMIC_EXECUTION]: The skill uses a Python script passed via the command line to parse JSON result files at runtime.\n
Evidence: python3 -c "import json,sys; d=json.load(open('playwright-results.json')); print(d.get('stats', d))" in SKILL.md.

e2e-testing