基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/huytieu/COG-second-brain --skill knowledge-consolidation命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Run one task through the V-model verification loop: CP-2 plan → CP-3 build → CP-3v component verify → CP-4 integration verify (full lane) → CP-5 acceptance. The worker never grades its own homework; evidence rows trace back to AC-n. Opt-in: invoke with /closed-loop or by asking for the closed loop, proper verification, or an evidence trail. Ordinary work does not run this.
Run a large, multi-session goal (e.g. shipping a whole side product) through the full V-model closed loop, one phase at a time, with cross-session state and a final north-star acceptance gate. Ultragoals never downgrade the lane: every phase runs CP-1→CP-6 with adversarial verification. Opt-in: invoke with /ultragoal or by calling something a long-running goal. Ordinary work does not run this.
Run one task through the V-model verification loop: CP-2 plan → CP-3 build → CP-3v component verify → CP-4 integration verify (full lane) → CP-5 acceptance. The worker never grades its own homework; evidence rows trace back to AC-n. Opt-in: invoke with /closed-loop or by asking for the closed loop, proper verification, or an evidence trail. Ordinary work does not run this.
| name | knowledge-consolidation |
| description | Build frameworks from scattered insights across all braindumps and notes |
| roles | ["all"] |
| integrations | [] |
Transform scattered insights from braindumps, daily briefs, and check-ins into coherent frameworks and "single source of truth" knowledge documents through pattern recognition and systematic synthesis.
Check agent_mode in 00-inbox/MY-PROFILE.md frontmatter:
agent_mode: team — delegate scanning and pattern extraction to parallel sub-agents (e.g., one per domain: personal braindumps, professional braindumps, project-specific content, daily briefs). Each agent identifies themes and patterns, then a synthesis agent combines findings into frameworks.agent_mode: solo (default) — handle all scanning, pattern recognition, and framework building directly. No delegation.Get current timestamp (REQUIRED before generating any files):
date '+%Y-%m-%d %H:%M' using Bash to get the actual current date and timecreated: frontmatter fielddate commandScan vault for unprocessed or partially processed content:
All braindumps since last consolidation:
02-personal/braindumps/03-professional/braindumps/04-projects/*/braindumps/00-inbox/braindump-*.md (mixed domain)Daily briefs and check-ins:
01-daily/briefs/01-daily/checkins/Any meeting transcripts or project documents in:
04-projects/*/planning/04-projects/*/resources/Determine scope:
status: "captured" or missing consolidation metadata)Gather statistics:
Apply systematic pattern detection across all content:
What comes up repeatedly?
What insights emerged together?
What patterns cross domains?
Where does thinking conflict?
Meta-patterns across all dimensions:
Synthesize patterns into actionable frameworks:
From scattered insights to fundamental truths:
Validate frameworks with source material:
When does framework apply/not apply?
How to use this framework:
Update and create knowledge base documents:
For each framework that needs updating, use the consolidated-knowledge template in references/templates.md.
Save to: 05-knowledge/consolidated/[framework-name]-framework.md
For newly identified frameworks, use the new-framework template in references/templates.md.
Save to: 05-knowledge/consolidated/[framework-name]-framework.md
Document the pattern using the pattern template in references/templates.md.
Save to: 05-knowledge/patterns/pattern-[name].md
Build the timeline entry using the thinking-evolution template in references/templates.md.
Save to: 05-knowledge/timeline/[topic]-evolution-YYYY-MM.md
Create the master consolidation document using the report template in references/templates.md.
Save to: 05-knowledge/consolidated/consolidation-YYYY-MM-DD.md
Mark processed braindumps: Update frontmatter in processed braindumps:
status: "consolidated"
consolidated_in: "[[consolidation-YYYY-MM-DD]]"
consolidated_date: "YYYY-MM-DD"
Archive outdated content:
Move superseded frameworks or insights to:
00-inbox/archive/[filename]-archived-YYYY-MM-DD.md
Add note explaining why archived and what supersedes it.
Maintain clean knowledge base:
After consolidation:
Consolidation is a loop-until-dry extraction with a completeness critic, not a single scan. See .claude/skills/loop-engineering/SKILL.md for the shared vocabulary.
The loop: scan a batch of in-scope documents → extract themes, patterns, and candidate framework principles → run the completeness critic ("any in-scope doc not yet read? any theme recurring across N+ docs that no framework captures yet?") → if the critic surfaces something new, run another extraction pass → stop when 2 passes in a row surface nothing new (dry). In agent_mode: team, the first scan fans out as one worker per domain (personal / professional / per-project / briefs); each returns its conclusions only, and a synthesis pass merges them.
The verifier (deterministic where it can be):
[[source]] is dropped, not published. This is mechanical and is COG's verification-first rule for consolidation.status: "consolidated" with a consolidated_in backlink.05-knowledge/consolidated/ so an existing framework is updated, not duplicated.Termination conditions (layered):
Patterns: loop-until-dry (the spine) + plan-execute-verify (each pass) + orchestrator-workers (team-mode domain scans) + completeness critic.
In-loop context: write the consolidation report incrementally and dedup new themes against the running set, not against the conversation. Externalizing to the report file is what keeps a large corpus from overflowing the window.
The knowledge consolidation skill embodies COG's self-evolving intelligence: