pi-skills/openspec/changes/archive/2026-06-10-goal-driven-loop/tasks.md

Tasks

• Create skill directory structure (.pi/skills/goal-driven-loop/)
• Write SKILL.md (≤500 words, Agent Skills standard compliant)
• Write references/error-handling.md (retry strategy, exception classification, stagnation rules, red flags)
• Create extension directory (.pi/extensions/goal-driven-loop/)
• Write trigger-detector.ts (trigger phrase matching + anti-trigger logic)
• Write state-injector.ts (GOAL.md parsing + injection text generation)
• Write ui-prompts.ts (timed auto-default confirms + notification helpers)
• Write index.ts (event wiring: input, before_agent_start)
• Add package.json for extension discovery
• Fix code block anti-trigger (strip code blocks before pattern matching)
• Fix quote block anti-trigger (only block if trigger exclusively in quotes)
• Fix blocker parsing (strip markdown list prefix)
• Add stagnation auto-recovery injection in extension
• Add comet fallback when openspec unavailable
• Replace any types with ExtensionContext
• Unit tests: 27/27 trigger-detector, 17/17 state-injector
• E2E tests: trigger, anti-trigger, abandon, stale state, state injection