MVPavan/via / .claude/skills/research
research
VerifiedUse when a decision needs multiple sources gathered and weighed. Answer a single factual lookup directly from an appropriate source.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 2.1 KB |
| agents/openai.yaml | config | 42 B |
| references/examples.md | markdown | 2.7 KB |
| references/frameworks.md | markdown | 3.1 KB |
| references/playbooks.md | markdown | 3.8 KB |
| references/source-selection.md | markdown | 2.2 KB |
| references/synthesis-engine.md | markdown | 1.9 KB |
references/synthesis-engine.md
View on GitHub# Synthesis Engine
Use relevant parts when a complex research question needs structured synthesis;
a short factual answer does not require this template.
## 1. Claim Table
List the claims that could materially change the answer.
For each claim record:
- Evidence.
- Source type.
- Date.
- Opposing evidence.
- Confidence.
- Why it matters.
## 2. Pattern Scan
Look for:
- Convergence: unrelated sources point to the same conclusion.
- Divergence: different segments, time periods, or incentives produce different answers.
- Silence: expected evidence is missing.
- Trend: direction changes over time.
- Bottleneck: one constraint determines whether the rest matters.
## 3. Fact To Insight
Turn facts into implications.
Formula:
`fact + context + implication for the user's decision`
If a paragraph does not answer "so what?", cut it or rewrite it.
## 4. Red Team
Before writing the conclusion, answer:
- What is the strongest case against this conclusion?
- What would make the opposite conclusion true?
- Are we seeing only winners, recent data, or vendor-friendly evidence?
- What source incentives could distort the picture?
- What failure mode would make this recommendation expensive or wrong?
Include the strongest counterargument in the document.
## 5. Confidence Calibration
For reports that benefit from explicit confidence labels:
- High: primary or authoritative sources agree, recent enough, and the red-team did not weaken it.
- Medium: evidence is credible but incomplete or partly indirect.
- Low: single source, narrow, anecdotal, stale, or conflicting evidence.
- Speculative: inference from patterns or analogies; state assumptions.
## 6. Final Narrative
Write in this order unless the topic demands otherwise:
1. Answer.
2. Scope and method.
3. Landscape or options.
4. Evidence and tradeoffs.
5. Counterarguments.
6. Decision points or recommendation.
7. Open questions and next research.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |