goal-backward-verification
Goal-Backward Verification
Verify that task outputs actually achieve stated goals through 4 progressive levels of verification. Works backward from the goal to verify each artifact exists, is substantive, is wired into the system, and passes functional tests.
When to Invoke
Skill({ skill: 'goal-backward-verification' });
Invoke when:
- A task claims completion and you need to verify the output
- QA agent is validating deliverables from a pipeline
- Post-pipeline drain gate needs artifact validation
- Reflection-agent is scoring task completeness
Verification Levels
More from oimiragieo/agent-studio
gcloud-cli
Google Cloud CLI operations and resource management
966pyqt6-ui-development-rules
PyQt6 desktop GUI development rules -- signal/slot architecture, QSS theming, QThread concurrency, layout management, and cross-platform rendering. Enforces MVC separation and responsive UI patterns.
575filesystem
File system operations guidance - read, write, search, and manage files using Claude Code's built-in tools.
361chrome-browser
Browser automation with two integrations - Chrome DevTools MCP (always available, performance tracing) and Claude-in-Chrome extension (authenticated sessions, GIF recording). Use DevTools for testing/debugging, Claude-in-Chrome for authenticated workflows.
304slack-notifications
Slack messaging, channels, and notifications - send messages, manage channels, interact with users, upload files, and add reactions. Use for team communication, incident notifications, and workflow alerts.
252context-compressor
Compress large context before reasoning to reduce token usage while preserving evidence. Use this whenever the user mentions huge files, long prompts, RAG payloads, prompt caching, expensive sessions, codebase context, chat history compaction, or wants the same answer quality with fewer tokens.
147