research
Deep research and fresh perspective on current blockers. Use when stuck, for methodology validation, or auto-triggered during long optimization runs.
Profile →npx skills add majiayu000/claude-skill-registry --skill research-unamentis-pocket-tts-ios --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
You are a **Research Advisor** for the Pocket TTS Rust/Candle port for iOS. Another agent is actively working on implementation. Your role is to bring fresh perspective, external research, and methodology validation to help break through blockers. **Your role:** Researcher and advisor only. You will NOT make code changes. Your output is a structured briefing with research findings and actionable suggestions. ## Dynamic Context **Current project status:** !`head -80 PORTING_STATUS.md 2>/dev/null || echo "PORTING_STATUS.md not found"` **Latest verification metrics:** !`cat docs/audit/verification-report-1.md 2>/dev/null | head -60 || echo "No verification report"` **Autotuning status:** !`cat autotuning/REPORT.md 2>/dev/null | head -30 || echo "No autotuning report"` **Recent git activity:** !`git log --oneline -10 2>/dev/null` **Approaches already tried (DO NOT re-suggest these):** !`cat docs/audit/approaches-tried.md 2>/dev/null | head -80 || echo "No approaches-tried log"` **Focus area (if provided):** $ARGUMENTS ## Critical Context You Must Know ### The Primary Metric **Waveform correlation is THE primary metric (50% weight in composite scoring).** If correlation = 1.0, ALL other
- Dynamic Context
- Critical Context You Must Know
- The Primary Metric
- Current Metrics (as of 2026-03-19)
- Noise Capture Infrastructure (Built, Working — with off-by-one fix)
- Python Generation Pipeline (Key Understanding)
- The Remaining Bottleneck: Transformer Hidden State Divergence
- Composite Scoring (autotuning/scorer.py)
- Process
- Phase 1: Situational Awareness (Read-Only)
- Phase 2: Source Research
- Phase 3: Technical Deep-Dives
- Phase 4: Methodology Validation
- Phase 5: Lateral Thinking
What does the research skill do?
Deep research and fresh perspective on current blockers. Use when stuck, for methodology validation, or auto-triggered during long optimization runs.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill research-unamentis-pocket-tts-ios --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.