AGENT SKILL

evidence-synthesis

This skill should be used when the user asks to "synthesize findings", "combine research", "cross-reference sources", "summarize evidence", "write a deep review", or mentions evidence synthesis, finding consolidation, cross-source analysis, or research summarization.

License 许可
MIT
最近更新
2026-08-03
Stars
★ 1
Forks
0

01 前言 Front Matter

name
evidence-synthesis
description
This skill should be used when the user asks to "synthesize findings", "combine research", "cross-reference sources", "summarize evidence", "write a deep review", or mentions evidence synthesis, finding consolidation, cross-source analysis, or research summarization.

03 SKILL.md 正文

Evidence Synthesis

Synthesis Workflow

  1. Gather — Read relevant YAML provenance records from docs/research/trail/findings/
  2. Group — Cluster findings by topic, tool, or theme
  3. Cross-reference — Identify agreements, contradictions, and gaps across sources
  4. Synthesize — Write deep review combining insights
  5. Score — Assess confidence level of synthesized conclusions

Progressive Disclosure Retrieval

Use 3-layer approach to avoid context bloat:

  • Layer 1 (index): file paths, titles, confidence — ~100 tokens/result
  • Layer 2 (context): finding summary, implication — ~500 tokens
  • Layer 3 (full detail): complete provenance record — ~500-1000 tokens

Search index FIRST, load context for relevant hits, full detail only when needed.

Deep Review Format

Write to docs/research/trail/deep-reviews/<topic>.md:

# <Topic> Deep Review

## Summary
<1-3 sentence executive summary>

## Key Findings
- Finding 1 (source: finding-<slug>.yaml, confidence: confirmed)
- Finding 2 (source: finding-<slug>.yaml, confidence: probable)

## Contradictions
<Where sources disagree and why>

## Gaps
<What remains unknown>

## Recommendations
<Actionable next steps>

Rules

  • Never load all findings into context at once — use progressive disclosure
  • Always cite source provenance records by filename
  • Flag contradictions explicitly — do not silently favor one source
  • Write synthesis to disk immediately, not at session end

04 同仓库更多 Skills

ray-manaloto
ray-manaloto/macos-development-environment

next-workflow-step

Use this skill when the user wants to transition to the next workflow step, wrap up a session cleanly, prepare a handoff, or says things like "next step", "continue the workflow", "wrap up", "hand off", "move to the next phase", "what's next in embrace", or "prepare for /clear". Also triggers when the current octo:embrace phase appears complete and no explicit next action is set.

Skill与智能体开发
★ 1
ray-manaloto
ray-manaloto/macos-development-environment

source-discovery

This skill should be used when the user asks to "find sources", "discover URLs", "catalog a source", "classify a URL", "search for libraries", "check what exists", or mentions source discovery, URL cataloging, source classification, or research source management.

★ 1
ray-manaloto
ray-manaloto/macos-development-environment

review-synthesis

This skill should be used when the user asks to "summarize review findings", "prioritize fixes", "rank review issues", "combine reviewer feedback", "create action items from review", or mentions review synthesis, finding prioritization, severity ranking, or review result summarization.

★ 1
ray-manaloto
ray-manaloto/macos-development-environment

research-scoring

This skill should be used when the user asks to "score research", "check research quality", "evaluate coverage", "measure improvement", "assess research completeness", or mentions research scoring, quality metrics, coverage gaps, improvement baseline, or research pipeline status.

资料检索与研究
★ 1
ray-manaloto
ray-manaloto/macos-development-environment

notebooklm-integration

This skill should be used when the user asks to "ingest into NotebookLM", "query NotebookLM", "add sources to notebook", "cross-source synthesis", "check stale sources", or mentions NotebookLM, notebook integration, source ingestion, or knowledge consolidation.

★ 1
ray-manaloto
ray-manaloto/macos-development-environment

debate-orchestration

This skill should be used when the user asks to "run a multi-model review", "debate code changes", "invoke reviewd", "get multiple model opinions", "run adversarial review", or mentions multi-model review, debate orchestration, reviewd CLI, codex review, gemini review, or adversarial code analysis.

★ 1