Skip to content

Self-evolving second brain with 35 AI skills, 10 agents, and people CRM. Closed-loop harness: a V-model verification lifecycle where the worker never grades its own homework. Plus paired anti-slop design skills for marketing and product UI. Works with Claude Code, Cursor, Kiro, Gemini CLI, Codex.

Add role packs and integration discovery (v3.3.0)

Huy Tieu584c8bc28336
diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
index 352d8b8..93ea519 100644
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@@ -1,6 +1,6 @@
{
"name": "cog-second-brain",
- "version": "3.2.0",
+ "version": "3.3.0",
"description": "A self-evolving second brain system using Claude Code, Obsidian markdown files, and Git version control. Features specialized AI agents for knowledge management, daily intelligence briefings, and strategic insights extraction.",
"author": {
"name": "Huy Tieu",
@@ -94,7 +94,8 @@
"onboardingTime": "2 minutes"
},
"directoryStructure": {
- ".claude/skills": "6 specialized agent skills (onboarding, braindump, daily-brief, weekly-checkin, knowledge-consolidation, url-dump)",
+ ".claude/skills": "10 specialized agent skills",
+ ".claude/roles": "7 role packs for personalized skill and integration recommendations",
"00-inbox": "Temporary processing area with profile and interests files",
"01-daily": "Daily briefs and check-ins",
"02-personal": "Personal domain (strict privacy)",
@@ -104,7 +105,8 @@
},
"skillArchitecture": {
"type": "agent-skills",
- "skills": 6,
+ "skills": 10,
+ "rolePacks": 7,
"invocation": "model-invoked",
"progressiveDisclosure": true,
"description": "Self-contained skills that combine orchestration and specialized processing using Claude's progressive disclosure pattern"
diff --git a/.claude/roles/_template.md b/.claude/roles/_template.md
new file mode 100644
index 0000000..6a68876
--- /dev/null
+++ b/.claude/roles/_template.md
@@ -0,0 +1,33 @@
+---
+role_id: my-role
+display_name: My Role
+aliases: []
+---
+
+# Role Pack: My Role
+
+> Copy this file, rename it to `your-role-id.md`, and customize the sections below.
+> The `role_id` must be lowercase with hyphens. `aliases` are alternative strings that match during onboarding (case-insensitive).
+
+## Recommended Skills
+
+*Order by relevance — most useful for this role first.*
+
+| Skill | Why it matters for this role |
+|---|---|
+| daily-brief | [Explain relevance] |
+| braindump | [Explain relevance] |
+
+## Recommended Integrations
+
+| Integration | Why it matters for this role |
+|---|---|
+| github | [Explain relevance] |
+
+## Suggested Agent Mode
+
+`solo` — *(or `team` if this role benefits from deeper, delegated analysis)*
+
+## Notes
+
+*Add any role-specific tips, workflow suggestions, or context here.*
diff --git a/.claude/roles/designer.md b/.claude/roles/designer.md
new file mode 100644
index 0000000..07064dc
--- /dev/null
+++ b/.claude/roles/designer.md
@@ -0,0 +1,31 @@
+---
+role_id: designer
+display_name: Designer
+aliases: [product designer, ux designer, ui designer, design lead, head of design, ux researcher, visual designer, interaction designer, design engineer]
+---
+
+# Role Pack: Designer
+
+## Recommended Skills
+
+| Skill | Why it matters for you |
+|---|---|
+| daily-brief | Design trends, tool updates, and UX research findings relevant to your focus areas |
+| braindump | Capture design critiques, user research observations, and creative ideas with automatic classification |
+| meeting-transcript | Extract design review decisions, feedback themes, and action items from meetings |
+| url-dump | Save design inspiration, case studies, and research articles with auto-extracted insights |
+| weekly-checkin | Spot patterns in design feedback, user research themes, and project momentum |
+| knowledge-consolidation | Build design systems documentation and UX pattern libraries from scattered notes |
+| onboarding | Initial setup — run this first |
+| update-cog | Keep your COG framework current |
+
+## Recommended Integrations
+
+| Integration | Why it matters for you |
+|---|---|
+| slack | Surface design feedback, stakeholder requests, and team discussions across channels |
+| notion | Pull from design specs, research repos, and team wikis for richer context |
+
+## Suggested Agent Mode
+
+`solo` — Designers typically work in focused creative sessions where direct interaction feels more natural.
diff --git a/.claude/roles/engineer.md b/.claude/roles/engineer.md
new file mode 100644
index 0000000..1d53b8a
--- /dev/null
+++ b/.claude/roles/engineer.md
@@ -0,0 +1,29 @@
+---
+role_id: engineer
+display_name: Engineer
+aliases: [software engineer, developer, dev, swe, frontend engineer, backend engineer, fullstack, full-stack, full stack engineer, software developer, programmer, coder]
+---
+
+# Role Pack: Engineer
+
+## Recommended Skills
+
+| Skill | Why it matters for you |
+|---|---|
+| daily-brief | Stay current on tools, frameworks, and security advisories relevant to your stack |
+| braindump | Capture technical learnings, debugging insights, and design ideas — auto-classified by domain |
+| url-dump | Save Stack Overflow answers, library docs, and blog posts with auto-extracted key takeaways |
+| weekly-checkin | Reflect on what you shipped, what you learned, and where you want to grow |
+| knowledge-consolidation | Build personal engineering reference docs from scattered notes and learnings |
+| onboarding | Initial setup — run this first |
+| update-cog | Keep your COG framework current |
+
+## Recommended Integrations
+
+| Integration | Why it matters for you |
+|---|---|
+| github | Track your PRs, review requests, and repository activity |
+
+## Suggested Agent Mode
+
+`solo` — Engineers typically have focused workflows where direct interaction is faster than delegated analysis.
diff --git a/.claude/roles/engineering-lead.md b/.claude/roles/engineering-lead.md
new file mode 100644
index 0000000..8d41f9f
--- /dev/null
+++ b/.claude/roles/engineering-lead.md
@@ -0,0 +1,35 @@
+---
+role_id: engineering-lead
+display_name: Engineering Lead
+aliases: [eng lead, engineering manager, em, tech lead, vp engineering, director of engineering, head of engineering, staff engineer, principal engineer]
+---
+
+# Role Pack: Engineering Lead
+
+## Recommended Skills
+
+| Skill | Why it matters for you |
+|---|---|
+| team-brief | Daily cross-reference of GitHub, Linear, Slack, and PostHog — see what shipped, what's stuck, and where to unblock your team |
+| comprehensive-analysis | Deep 7-day analysis for sprint retros, tech debt reviews, and engineering health assessments |
+| meeting-transcript | Capture architecture decisions, sprint planning outcomes, and 1:1 action items with structured ownership tracking |
+| daily-brief | Stay current on engineering trends, security advisories, and technology shifts relevant to your stack |
+| braindump | Capture technical design thoughts, incident learnings, and team observations with automatic classification |
+| weekly-checkin | Spot patterns across sprints — recurring blockers, velocity trends, team dynamics shifts |
+| knowledge-consolidation | Build engineering playbooks and decision records from scattered technical notes |
+| url-dump | Save technical articles, architecture references, and tool evaluations with auto-extracted insights |
+| onboarding | Initial setup — run this first |
+| update-cog | Keep your COG framework current |
+
+## Recommended Integrations
+
+| Integration | Why it matters for you |
+|---|---|
+| github | PRs, commits, reviews, CI/CD status — the core data source for engineering intelligence |
+| linear | Sprint tracking, issue status, blockers — enables two-way sync from team-brief back to Linear |
+| slack | Team discussions, incident threads, and engineering channel activity inform team-brief context |
+| posthog | Feature flag status, error rates, and performance metrics — the production health layer |
+
+## Suggested Agent Mode
+
+`team` — Engineering leads manage enough data streams that delegated analysis produces significantly better results, especially for comprehensive-analysis and team-brief.
diff --git a/.claude/roles/founder.md b/.claude/roles/founder.md
new file mode 100644
index 0000000..aa48069
--- /dev/null
+++ b/.claude/roles/founder.md
@@ -0,0 +1,37 @@
+---
+role_id: founder
+display_name: Founder
+aliases: [ceo, co-founder, cofounder, startup founder, cto, coo, chief executive, entrepreneur, solo founder, technical founder]
+---
+
+# Role Pack: Founder
+
+## Recommended Skills
+
+| Skill | Why it matters for you |
+|---|---|
+| team-brief | Daily cross-reference of GitHub, Linear, Slack, and PostHog — full visibility across your entire team's output without chasing updates |
+| comprehensive-analysis | Deep 7-day analysis for board prep, investor updates, and strategic planning — pulls from every data source |
+| meeting-transcript | Capture investor meeting commitments, board decisions, and team alignment outcomes with structured ownership |
+| daily-brief | Market intelligence, competitor moves, fundraising news, and industry trends — personalized to your space |
+| braindump | Capture strategic pivots, investor feedback, and cross-functional observations with automatic domain classification |
+| weekly-checkin | Cross-domain pattern analysis — connects product progress to market timing to team dynamics |
+| knowledge-consolidation | Build strategic frameworks, pitch narratives, and decision logs from scattered founder thoughts |
+| url-dump | Save competitor launches, market analyses, and investor content with auto-extracted insights |
+| onboarding | Initial setup — run this first |
+| update-cog | Keep your COG framework current |
+
+## Recommended Integrations
+
+| Integration | Why it matters for you |
+|---|---|
+| github | Engineering velocity, release cadence, and technical health — essential for founder visibility |
+| linear | Sprint progress, team blockers, and priority alignment — enables two-way sync from team-brief |
+| slack | Cross-functional discussions, team sentiment, and decision trails across all channels |
+| posthog | Product-market fit signals — activation funnels, retention curves, and feature adoption |
+| notion | Company wiki, investor updates, and strategic documents for richer analysis context |
+| hackmd | Collaborative notes from team meetings and strategy sessions |
+
+## Suggested Agent Mode
+
+`team` — Founders juggle the most context. Team mode delegates research across data sources and produces the comprehensive analysis founders need for decision-making.
diff --git a/.claude/roles/marketer.md b/.claude/roles/marketer.md
new file mode 100644
index 0000000..196684c
--- /dev/null
+++ b/.claude/roles/marketer.md
@@ -0,0 +1,31 @@
+---
+role_id: marketer
+display_name: Marketer
+aliases: [marketing manager, growth marketer, content marketer, head of marketing, marketing lead, vp marketing, director of marketing, growth lead, demand gen, content strategist, brand manager]
+---
+
+# Role Pack: Marketer
+
+## Recommended Skills
+
+| Skill | Why it matters for you |
+|---|---|
+| daily-brief | Industry news, competitor campaigns, and market trends relevant to your channels and audience |
+| braindump | Capture campaign ideas, audience insights, and content angles with automatic domain classification |
+| url-dump | Save competitor content, industry reports, and campaign inspiration with auto-extracted key takeaways |
+| knowledge-consolidation | Build messaging frameworks, audience personas, and campaign playbooks from scattered marketing notes |
+| weekly-checkin | Spot patterns in campaign performance, content themes, and market shifts |
+| onboarding | Initial setup — run this first |
+| update-cog | Keep your COG framework current |
+
+## Recommended Integrations
+
+| Integration | Why it matters for you |
+|---|---|
+| slack | Surface campaign discussions, cross-team requests, and customer feedback from team channels |
+| notion | Pull from content calendars, brand guidelines, and marketing wikis for richer context |
+| posthog | Funnel metrics, conversion rates, and campaign attribution — the data layer for growth marketing |
+
+## Suggested Agent Mode
+
+`solo` — Marketers typically work in focused creative and analytical sessions where direct interaction is most efficient.
diff --git a/.claude/roles/product-manager.md b/.claude/roles/product-manager.md
new file mode 100644
index 0000000..c550070
--- /dev/null
+++ b/.claude/roles/product-manager.md
@@ -0,0 +1,37 @@
+---
+role_id: product-manager
+display_name: Product Manager
+aliases: [pm, product lead, product owner, head of product, vp product, director of product, product person]
+---
+
+# Role Pack: Product Manager
+
+## Recommended Skills
+
+| Skill | Why it matters for you |
+|---|---|
+| team-brief | Daily cross-reference of GitHub, Linear, Slack, and PostHog — see what shipped, what's blocked, and what needs your attention without chasing updates |
+| comprehensive-analysis | Deep 7-day analysis for weekly reviews, board prep, and stakeholder updates — pulls from every data source |
+| meeting-transcript | Turn meeting recordings into structured decisions, action items, and ownership — never lose track of commitments |
+| daily-brief | Personalized news intelligence on your industry, competitors, and market trends |
+| braindump | Capture product ideas, user feedback patterns, and strategic thoughts with automatic domain classification |
+| weekly-checkin | Cross-domain pattern analysis — connects product metrics to team health to market shifts |
+| knowledge-consolidation | Build product strategy frameworks from scattered insights across braindumps and notes |
+| url-dump | Save competitor pages, user research articles, and market reports with auto-extracted insights |
+| onboarding | Initial setup — run this first |
+| update-cog | Keep your COG framework current |
+
+## Recommended Integrations
+
+| Integration | Why it matters for you |
+|---|---|
+| github | Track PRs, releases, and engineering velocity — feeds into team-brief and comprehensive-analysis |
+| linear | See issue status, sprint progress, and blockers — enables two-way sync from team-brief back to Linear |
+| slack | Surface team discussions, decisions made in channels, and sentiment — critical for team-brief context |
+| posthog | Product analytics — feature adoption, funnel metrics, and usage trends inform your briefs |
+| notion | Pull from product specs, PRDs, and team wikis for richer analysis context |
+| hackmd | Access shared meeting notes and collaborative documents |
+
+## Suggested Agent Mode
+
+`team` — PMs benefit from delegated deep analysis. Team mode sends research and data gathering to specialist sub-agents, producing more thorough briefs and analyses.
diff --git a/.claude/skills/braindump/SKILL.md b/.claude/skills/braindump/SKILL.md
index 598f9d4..8557505 100644
--- a/.claude/skills/braindump/SKILL.md
+++ b/.claude/skills/braindump/SKILL.md
@@ -1,6 +1,8 @@
---
name: braindump
description: Quick capture of raw thoughts with intelligent domain classification and competitive intelligence extraction
+roles: [all]
+integrations: []
---
# COG Braindump Skill
diff --git a/.claude/skills/comprehensive-analysis/SKILL.md b/.claude/skills/comprehensive-analysis/SKILL.md
index b5c4f8f..d244d67 100644
--- a/.claude/skills/comprehensive-analysis/SKILL.md
+++ b/.claude/skills/comprehensive-analysis/SKILL.md
@@ -1,6 +1,8 @@
---
name: comprehensive-analysis
description: Deep-dive 7-day analysis across all data sources for weekly reviews, board prep, and strategic planning
+roles: [product-manager, engineering-lead, founder]
+integrations: [github, linear, slack, posthog]
---
# COG Comprehensive Analysis Skill
diff --git a/.claude/skills/daily-brief/SKILL.md b/.claude/skills/daily-brief/SKILL.md
index a8d872b..ffc4c02 100644
--- a/.claude/skills/daily-brief/SKILL.md
+++ b/.claude/skills/daily-brief/SKILL.md
@@ -1,6 +1,8 @@
---
name: daily-brief
description: Generate personalized news intelligence with verified sources (7-day freshness requirement)
+roles: [all]
+integrations: [web-search]
---
# COG Daily Brief Skill
diff --git a/.claude/skills/knowledge-consolidation/SKILL.md b/.claude/skills/knowledge-consolidation/SKILL.md
index 732a9cd..66cabff 100644
--- a/.claude/skills/knowledge-consolidation/SKILL.md
+++ b/.claude/skills/knowledge-consolidation/SKILL.md
@@ -1,6 +1,8 @@
---
name: knowledge-consolidation
description: Build frameworks from scattered insights across all braindumps and notes
+roles: [all]
+integrations: []
---
# COG Knowledge Consolidation Skill
diff --git a/.claude/skills/meeting-transcript/SKILL.md b/.claude/skills/meeting-transcript/SKILL.md
index 9cc0ddf..1d189ee 100644
--- a/.claude/skills/meeting-transcript/SKILL.md
+++ b/.claude/skills/meeting-transcript/SKILL.md
@@ -1,6 +1,8 @@
---
name: meeting-transcript
description: Process meeting recordings and notes into structured decisions, action items, and team dynamics with intelligent noise filtering
+roles: [product-manager, engineering-lead, founder, designer]
+integrations: []
---
# COG Meeting Transcript Skill
diff --git a/.claude/skills/onboarding/SKILL.md b/.claude/skills/onboarding/SKILL.md
index f9ee357..5803157 100644
--- a/.claude/skills/onboarding/SKILL.md
+++ b/.claude/skills/onboarding/SKILL.md
@@ -1,6 +1,8 @@
---
name: onboarding
description: Personalize COG for your workflow - creates profile, interests, and watchlist files with guided setup (run this first!)
+roles: [all]
+integrations: []
---
# COG Onboarding Skill
@@ -96,6 +98,83 @@ Which do you prefer? (Solo is great for most people - team mode is for power use
**Wait for confirmation**, then generate everything. If they say "looks good" or similar, proceed. If they correct something, update and proceed without re-confirming. Default to `solo` if they don't express a preference.
+### 5.5. Role Pack Matching
+
+After extracting the user's role text in Step 3, scan `.claude/roles/*.md` for a matching role pack:
+
+1. Read each role pack file's YAML frontmatter (`role_id` and `aliases`)
+2. Compare the user's extracted role text (case-insensitive) against:
+ - Exact `role_id` match (e.g., "product-manager")
+ - Any string in `aliases` (e.g., "pm", "product lead", "head of product")
+ - Fuzzy substring match (e.g., "product manager at a fintech startup" contains "product manager")
+3. If a match is found:
+ - Store the matched `role_id` as `role_pack` in the MY-PROFILE.md frontmatter
+ - Present role-specific recommendations:
+ ```
+ As a [Role Display Name], here are the skills and integrations that'll be most useful for you:
+
+ **Recommended skills** (ordered by relevance for your role):
+ [List top 5-6 skills from the role pack with the "Why it matters for you" context]
+
+ **Recommended integrations**:
+ [List integrations from the role pack with role-specific explanations]
+ ```
+ - Use the role pack's suggested `agent_mode` as the default (instead of `solo`)
+4. If no match is found:
+ - Set `role_pack: custom` in MY-PROFILE.md
+ - Recommend core skills only (those with `roles: [all]`)
+ - Ask about common integrations: "Do you use GitHub, Slack, or any other tools you'd like COG to connect with?"
+
+### 5.6. Integration Discovery
+
+After role pack matching, set up the user's integration preferences:
+
+1. If a role pack was matched, present its recommended integrations with role-specific context:
+ ```
+ Based on your role, these integrations would give COG the most context:
+
+ [For each integration in role pack:]
+ - **[Integration]** — [Why it matters for you, from role pack]
+
+ Which of these do you already use? And are there any other tools you'd like to connect?
+ ```
+
+2. Parse the user's response:
+ - Services they confirm using → add to **Active** section of MY-INTEGRATIONS.md
+ - Services they don't mention or say no to → add to **Disabled** section
+ - Additional services they mention → add to **Active** section
+ - Always add `ElevenLabs` to **Disabled** unless explicitly requested
+
+3. Generate `00-inbox/MY-INTEGRATIONS.md`:
+ ```markdown
+ ---
+ type: integrations
+ created: YYYY-MM-DD
+ tags: ["#integrations", "#config", "#cog"]
+ ---
+
+ # My Integrations
+
+ *COG checks this file before using any external service. Edit anytime.*
+
+ ## Active
+ [For each confirmed integration:]
+ - **[Service]**: [Brief description of how COG uses it]
+
+ ## Disabled
+ [For each declined/unmentioned integration:]
+ - **[Service]**: Skipped during onboarding. Enable anytime by moving to Active section.
+
+ ---
+
+ *Move services between Active and Disabled sections to control what COG connects to.*
+ ```
+
+4. If no role pack was matched, ask about common integrations conversationally:
+ ```
+ COG can connect with tools like GitHub, Slack, Linear, Notion, and PostHog for richer analysis. Do you use any of these? (Totally optional - COG works great without them too.)
+ ```
+
### 6. Generate Profile Documents
Create the following markdown files:
@@ -106,7 +185,8 @@ Create the following markdown files:
type: profile
created: YYYY-MM-DD
onboarding_completed: true
-agent_mode: solo
+role_pack: [matched role_id or "custom"]
+agent_mode: [solo or team, based on role pack suggestion]
tags: ["#profile", "#config", "#cog"]
---
@@ -115,10 +195,11 @@ tags: ["#profile", "#config", "#cog"]
## About Me
- **Name**: [Name]
- **Role**: [Job/role/main activity]
+- **Role Pack**: [Display name from matched role pack, or "Custom" if no match]
- **Profile Created**: [Date]
## Settings
-- **Agent Mode**: solo *(solo = handle everything directly; team = delegate to specialist sub-agents for deeper results)*
+- **Agent Mode**: [solo/team] *(solo = handle everything directly; team = delegate to specialist sub-agents for deeper results)*
## Active Projects
[If they mentioned projects:]
@@ -293,13 +374,45 @@ Your COG is now personalized and ready to use. Here's how to get started:
I've created these documents to store your preferences:
-- **[[MY-PROFILE]]** - Your basic info and workflow preferences
+- **[[MY-PROFILE]]** - Your basic info, role pack, and workflow preferences
- **[[MY-INTERESTS]]** - Topics for your daily briefs
+- **[[MY-INTEGRATIONS]]** - Your active and disabled integrations
- **[[03-professional/COMPETITIVE-WATCHLIST]]** - Companies you're tracking *(if applicable)*
**You can edit these files anytime.** COG reads them when you use skills, so your changes take effect immediately.
-## Quick Start Skills
+## Skills for Your Role
+
+[If role pack was matched:]
+As a **[Role Display Name]**, these skills are ordered by relevance for you:
+
+[List skills from role pack in order, with brief "why it matters" from the role pack. Format as:]
+1. **[skill-name]** — [Role-specific explanation]
+2. **[skill-name]** — [Role-specific explanation]
+[...continue for all recommended skills]
+
+[If no role pack match:]
+Here are COG's core skills available to everyone:
+
+1. **daily-brief** — Personalized news intelligence
+2. **braindump** — Capture and classify thoughts
+3. **weekly-checkin** — Weekly pattern analysis
+4. **knowledge-consolidation** — Build frameworks from scattered notes
+5. **url-dump** — Save URLs with auto-extracted insights
+6. **update-cog** — Keep COG framework current
+
+## Your Integrations
+
+[If integrations were configured:]
+**Active**: [List active integrations]
+**Disabled**: [List disabled integrations]
+
+You can change these anytime by editing [[MY-INTEGRATIONS]].
+
+[If no integrations configured:]
+No integrations configured yet. COG works great standalone — add integrations anytime by editing `00-inbox/MY-INTEGRATIONS.md`.
+
+## Quick Start
### 1. Daily Morning Routine
Invoke the daily-brief skill to get your personalized intelligence briefing covering:
@@ -392,11 +505,13 @@ Then intelligently handle whatever they say - whether it's adding projects, chan
## Success Criteria
Onboarding is successful when:
-1. `MY-PROFILE.md` created in `00-inbox/`
+1. `MY-PROFILE.md` created in `00-inbox/` with `role_pack` in frontmatter
2. `MY-INTERESTS.md` created in `00-inbox/`
-3. Project directories and overviews created (if applicable)
-4. `WELCOME-TO-COG.md` guide created
-5. User understands next steps and where their profile is stored
+3. `MY-INTEGRATIONS.md` created in `00-inbox/` with active/disabled sections
+4. Role pack matched (or set to `custom`) and recommendations presented
+5. Project directories and overviews created (if applicable)
+6. `WELCOME-TO-COG.md` guide created with role-specific skill ordering
+7. User understands next steps and where their profile is stored
## Error Handling
@@ -418,8 +533,9 @@ Onboarding is successful when:
## Privacy & Data
All configuration data is stored as markdown files in:
-- `00-inbox/MY-PROFILE.md` - Basic profile
+- `00-inbox/MY-PROFILE.md` - Basic profile with role pack
- `00-inbox/MY-INTERESTS.md` - Interest areas
+- `00-inbox/MY-INTEGRATIONS.md` - Active/disabled external service integrations
- `03-professional/COMPETITIVE-WATCHLIST.md` - Competitive tracking
- `04-projects/[project]/PROJECT-OVERVIEW.md` - Project details
diff --git a/.claude/skills/team-brief/SKILL.md b/.claude/skills/team-brief/SKILL.md
index fb7e8e3..1e2ee9b 100644
--- a/.claude/skills/team-brief/SKILL.md
+++ b/.claude/skills/team-brief/SKILL.md
@@ -1,6 +1,8 @@
---
name: team-brief
description: Generate daily team intelligence brief by cross-referencing GitHub, Linear, Slack, PostHog, meetings, and braindumps with two-way Linear sync-back
+roles: [product-manager, engineering-lead, founder]
+integrations: [github, linear, slack, posthog, hackmd]
---
# COG Team Intelligence Brief Skill
diff --git a/.claude/skills/update-cog/SKILL.md b/.claude/skills/update-cog/SKILL.md
index 7033202..f3f62bd 100644
--- a/.claude/skills/update-cog/SKILL.md
+++ b/.claude/skills/update-cog/SKILL.md
@@ -1,6 +1,8 @@
---
name: update-cog
description: Check for and apply upstream COG framework updates (skills, docs, scripts) without touching personal content
+roles: [all]
+integrations: [git]
---
# COG Update Skill
diff --git a/.claude/skills/url-dump/SKILL.md b/.claude/skills/url-dump/SKILL.md
index bf30997..e0a93b3 100644
--- a/.claude/skills/url-dump/SKILL.md
+++ b/.claude/skills/url-dump/SKILL.md
@@ -1,6 +1,8 @@
---
name: url-dump
description: Quick capture URLs with automatic content extraction, insights, and categorization into knowledge booklets
+roles: [all]
+integrations: [web-fetch]
---
# COG URL Dump Skill
diff --git a/.claude/skills/weekly-checkin/SKILL.md b/.claude/skills/weekly-checkin/SKILL.md
index b543079..898a040 100644
--- a/.claude/skills/weekly-checkin/SKILL.md
+++ b/.claude/skills/weekly-checkin/SKILL.md
@@ -1,6 +1,8 @@
---
name: weekly-checkin
description: Cross-domain pattern analysis and strategic reflection for weekly review
+roles: [all]
+integrations: []
---
# COG Weekly Check-in Skill
diff --git a/AGENTS.md b/AGENTS.md
index ddebb77..bd49c77 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -27,10 +27,13 @@ This document defines the available commands/skills for AI agents interacting wi
3. Only asks a follow-up if required info (name, role, interests) is still missing
4. Asks about agent mode preference (solo vs team) during confirmation
5. Confirms extracted info before creating files
-6. Creates `00-inbox/MY-PROFILE.md` with basic preferences and agent_mode setting
-7. Creates `00-inbox/MY-INTERESTS.md` with topics for daily briefs
-8. Optionally creates project structures in `04-projects/` and `03-professional/COMPETITIVE-WATCHLIST.md` (only if mentioned)
-9. Generates a welcome guide
+6. Matches role to a role pack (`.claude/roles/*.md`) for personalized skill and integration recommendations
+7. Discovers integrations — presents role-specific recommendations, asks which tools the user already uses
+8. Creates `00-inbox/MY-PROFILE.md` with role_pack, agent_mode, and preferences
+9. Creates `00-inbox/MY-INTERESTS.md` with topics for daily briefs
+10. Creates `00-inbox/MY-INTEGRATIONS.md` with active/disabled integrations
+11. Optionally creates project structures in `04-projects/` and `03-professional/COMPETITIVE-WATCHLIST.md` (only if mentioned)
+12. Generates a welcome guide with role-ordered skills and integration status
**Agent modes:**
- **Solo** (default): All skills handle everything directly in one conversation
@@ -245,9 +248,11 @@ This document defines the available commands/skills for AI agents interacting wi
```
COG-second-brain/
+├── .claude/roles/ # Role packs for personalized recommendations
├── 00-inbox/ # Landing zone, profile files
-│ ├── MY-PROFILE.md # User profile (created by onboarding)
-│ └── MY-INTERESTS.md # User interests (created by onboarding)
+│ ├── MY-PROFILE.md # User profile with role pack (created by onboarding)
+│ ├── MY-INTERESTS.md # User interests (created by onboarding)
+│ └── MY-INTEGRATIONS.md # Active/disabled integrations (created by onboarding)
├── 01-daily/ # Daily content
│ ├── briefs/ # Daily intelligence briefs
│ └── checkins/ # Weekly check-ins
@@ -286,15 +291,25 @@ COG-second-brain/
## Configuration
All configuration is stored as readable markdown files:
-- `00-inbox/MY-PROFILE.md` - Basic profile and active projects
+- `00-inbox/MY-PROFILE.md` - Profile, role pack, agent mode, and active projects
- `00-inbox/MY-INTERESTS.md` - Topics for news curation
+- `00-inbox/MY-INTEGRATIONS.md` - Active/disabled external service integrations
- `03-professional/COMPETITIVE-WATCHLIST.md` - Companies/people to track
Edit these files anytime - changes take effect immediately.
+### Role Packs
+
+COG matches your role to a role pack during onboarding. Role packs (in `.claude/roles/`) define:
+- Which skills are most relevant for your role
+- Which integrations to recommend
+- Suggested agent mode (solo vs team)
+
+Available packs: Product Manager, Engineering Lead, Engineer, Designer, Founder, Marketer. Create custom packs from `_template.md`.
+
## Version & Updates
-COG tracks its version in `COG-VERSION` (currently 3.2.0). To check for updates:
+COG tracks its version in `COG-VERSION` (currently 3.3.0). To check for updates:
- Run `/update-cog` in any supported agent
- Or use the shell script: `./cog-update.sh --check`
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 356219e..1744950 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,6 +2,64 @@
All notable changes to COG (Cognition + Obsidian + Git) will be documented in this file.
+## [3.3.0] - 2026-02-25
+
+### Role Packs & Integration Discovery
+
+COG now matches your role during onboarding to personalize skill recommendations and integration suggestions. A PM and an engineer see different skill priorities. New roles can be added by dropping a file.
+
+### Added
+
+#### Role Packs (`.claude/roles/`)
+- **7 role pack files** — each defines per-role skill recommendations and integration needs:
+ - `_template.md` — starter for custom roles
+ - `product-manager.md` — skills: team-brief, comprehensive-analysis, meeting-transcript, daily-brief, braindump, etc. Integrations: GitHub, Linear, Slack, PostHog, Notion, HackMD
+ - `engineering-lead.md` — engineering-focused with team management emphasis. Integrations: GitHub, Linear, Slack, PostHog
+ - `engineer.md` — individual contributor focus. Integrations: GitHub
+ - `designer.md` — design and UX research focus. Integrations: Slack, Notion
+ - `founder.md` — all skills, all integrations recommended
+ - `marketer.md` — growth and content focus. Integrations: Slack, Notion, PostHog
+- Each role pack contains YAML frontmatter with `role_id`, `display_name`, and `aliases` for fuzzy matching during onboarding
+
+#### Onboarding Enhancements
+- **Step 5.5 — Role Pack Matching**: After extracting role text, scans `.claude/roles/*.md` for matching `role_id` or `aliases`. Presents role-specific skill and integration recommendations
+- **Step 5.6 — Integration Discovery**: Presents role pack's recommended integrations with role-specific explanations. Generates `00-inbox/MY-INTEGRATIONS.md` with Active/Disabled sections
+- **Updated MY-PROFILE.md template**: Now includes `role_pack` in YAML frontmatter
+- **Updated WELCOME-TO-COG.md template**: New "Skills for Your Role" section with role-ordered skills, and "Your Integrations" section
+
+#### Skill Metadata
+- All 10 skills now have `roles` and `integrations` fields in YAML frontmatter
+ - Core skills (onboarding, braindump, daily-brief, etc.): `roles: [all]`
+ - Team skills (team-brief, comprehensive-analysis): `roles: [product-manager, engineering-lead, founder]`
+ - Meeting-transcript: `roles: [product-manager, engineering-lead, founder, designer]`
+
+#### Framework Configuration
+- **`CLAUDE.md`** rewritten as universal framework file with Role Packs and Integration Preferences sections (no longer personal config)
+
+### Changed
+
+#### Documentation
+- **`README.md`** — Added "Role Packs" section, updated vault structure diagram with `.claude/roles/`, updated roadmap
+- **`SETUP.md`** — Updated skill counts, added role packs to onboarding output, added MY-INTEGRATIONS.md
+- **`AGENTS.md`** — Updated onboarding command with role matching and integration discovery, added team intelligence skills, updated vault structure and configuration section
+- **`GEMINI.md`** — Updated vault structure with role packs and integrations file
+- **`CONTRIBUTING.md`** — Updated skill frontmatter convention to include `roles` and `integrations` fields, added role pack contribution guidelines
+- **`CHANGELOG.md`** — This entry
+
+#### Version & Metadata
+- **`COG-VERSION`** — Bumped 3.2.0 → 3.3.0
+- **`.claude-plugin/plugin.json`** — Bumped version, updated skill count to 10, added `rolePacks: 7` metadata
+- **`marketplace-entry.json`** — Bumped version to 3.3.0
+- **`cog-update.sh`** — Added 7 role pack files, CLAUDE.md, and 3 team intelligence skills to FRAMEWORK_FILES array
+
+### Design Decisions
+- **File-based role packs**: New roles added by dropping a `.md` file — no code changes needed
+- **Alias-based matching**: Fuzzy matching via aliases handles variations like "PM", "product lead", "head of product"
+- **Integration discovery during onboarding**: Rather than skills failing at runtime, integrations are configured upfront
+- **CLAUDE.md as framework file**: Moves from personal config to universal instructions that work for any user
+
+---
+
## [3.2.0] - 2026-02-09
### Upstream Update System
diff --git a/CLAUDE.md b/CLAUDE.md
new file mode 100644
index 0000000..8059e46
--- /dev/null
+++ b/CLAUDE.md
@@ -0,0 +1,51 @@
+# COG Second Brain — Framework Instructions
+
+## Integration Preferences
+
+Before using any external integration in a skill, check `00-inbox/MY-INTEGRATIONS.md`:
+
+- **Active integrations**: Use normally.
+- **Disabled integrations**: Skip silently. Do not attempt to call their tools, do not suggest setting them up, do not mention them in output.
+- **Unknown integrations** (not listed in either section): Ask the user if they want to set it up. If they say no, add it to the Disabled section.
+
+## Role Packs
+
+COG uses role packs (`.claude/roles/*.md`) to personalize skill recommendations and integration suggestions per user role.
+
+### How role matching works
+1. During onboarding, the user's role text is matched against `role_id` and `aliases` in each role pack's YAML frontmatter.
+2. The matched role pack is stored as `role_pack` in `00-inbox/MY-PROFILE.md` frontmatter.
+3. When suggesting skills or workflows, check the user's `role_pack` and order recommendations by role relevance.
+
+### Role-aware behavior
+- **Skill suggestions**: When a user asks "what can COG do?" or similar, prioritize skills listed in their role pack. Show role-specific explanations from the pack.
+- **Integration prompts**: When a skill needs an integration the user hasn't set up, check their role pack to provide role-specific context for why it matters.
+- **No role pack match**: If the user's role doesn't match any pack, recommend core skills (`roles: [all]`) and let them discover team skills organically.
+
+### Available role packs
+Role packs live in `.claude/roles/`. New roles can be added by dropping a file following the `_template.md` format.
+
+## Vault Structure
+
+### User configuration files (`00-inbox/`)
+- `MY-PROFILE.md` — User info, role pack, agent mode, active projects
+- `MY-INTERESTS.md` — Topics for daily briefs
+- `MY-INTEGRATIONS.md` — Active/disabled external service integrations
+
+### Professional tracking (`03-professional/`)
+- `COMPETITIVE-WATCHLIST.md` — Companies/people being tracked
+
+### Framework files (updated via `cog-update.sh` or `/update-cog`)
+- `.claude/skills/` — Claude Code skills (10 skills)
+- `.claude/roles/` — Role packs for personalized recommendations
+- `.kiro/powers/` — Kiro powers
+- `.gemini/commands/` — Gemini CLI commands
+- `AGENTS.md` — Universal agent documentation
+
+### Content directories (never touched by updates)
+- `00-inbox/` — Profiles, interests, integrations
+- `01-daily/` — Briefs and check-ins
+- `02-personal/` — Personal braindumps (private)
+- `03-professional/` — Professional braindumps and strategy
+- `04-projects/` — Per-project tracking
+- `05-knowledge/` — Consolidated insights and patterns
diff --git a/COG-VERSION b/COG-VERSION
index 944880f..15a2799 100644
--- a/COG-VERSION
+++ b/COG-VERSION
@@ -1 +1 @@
-3.2.0
+3.3.0
diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index 6b626cf..903a3c1 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -45,9 +45,10 @@ We welcome ideas for new features or improvements:
#### Making Changes
1. **Skills**: New skills should be added in all supported formats:
- - `.claude/skills/[name]/SKILL.md` - Claude Code format
+ - `.claude/skills/[name]/SKILL.md` - Claude Code format (include `roles` and `integrations` in frontmatter)
- `.kiro/powers/cog-[name]/POWER.md` - Kiro format
- Update `agents.md` - Universal documentation
+ - If the skill is role-specific, add it to relevant role packs in `.claude/roles/`
2. **Templates**: Follow existing YAML frontmatter conventions
3. **Documentation**: Update README.md if adding major features
4. **Examples**: Provide examples of your feature in action
@@ -116,6 +117,8 @@ Before submitting:
---
name: skill-name
description: What this skill does
+roles: [all]
+integrations: [github, slack]
---
# Skill Name
diff --git a/GEMINI.md b/GEMINI.md
index ee3c417..3a74e18 100644
--- a/GEMINI.md
+++ b/GEMINI.md
@@ -9,7 +9,7 @@ You are the user's personal knowledge agent. Help them capture thoughts, stay in
## Vault Structure
```
-00-inbox/ → Landing zone, profile files (MY-PROFILE.md, MY-INTERESTS.md)
+00-inbox/ → Landing zone, profile files (MY-PROFILE.md, MY-INTERESTS.md, MY-INTEGRATIONS.md)
01-daily/ → briefs/, checkins/
02-personal/ → braindumps/
03-professional/ → braindumps/, COMPETITIVE-WATCHLIST.md
@@ -20,8 +20,9 @@ You are the user's personal knowledge agent. Help them capture thoughts, stay in
## User Profile
-Read `00-inbox/MY-PROFILE.md` for user name, role, and active projects.
+Read `00-inbox/MY-PROFILE.md` for user name, role, role pack, and active projects.
Read `00-inbox/MY-INTERESTS.md` for topics and preferred news sources.
+Read `00-inbox/MY-INTEGRATIONS.md` for active/disabled external service integrations.
Read `03-professional/COMPETITIVE-WATCHLIST.md` for companies to track.
## Available Skills
diff --git a/README.md b/README.md
index f10fe0c..92275fc 100644
--- a/README.md
+++ b/README.md
@@ -60,6 +60,10 @@ Done — COG is personalized and ready in ~2 minutes. See [SETUP.md](SETUP.md) f
| **meeting-transcript** | Process meeting recordings into structured decisions, action items, and team dynamics | "Process this meeting" |
| **comprehensive-analysis** | Deep 7-day analysis for weekly reviews, board prep, or strategic planning (~8-12 min) | "Weekly analysis" / "Board prep" |
+### Role Packs (Personalized Recommendations)
+
+COG matches your role during onboarding to a **role pack** that prioritizes the most relevant skills and integrations for you. Available role packs: Product Manager, Engineering Lead, Engineer, Designer, Founder, Marketer — or create your own from the template.
+
> **New to team skills?** These require GitHub CLI (`gh`) and work best with Linear, Slack, and PostHog MCP integrations. They degrade gracefully — start with just GitHub and add integrations over time. See [SETUP.md](SETUP.md) for configuration.
## The Evolution Cycle
@@ -96,11 +100,13 @@ graph TD
```
COG-second-brain/
-├── .claude/skills/ # Claude Code skills
+├── .claude/skills/ # Claude Code skills (10)
+├── .claude/roles/ # Role packs (7) — personalized recommendations
├── .kiro/powers/ # Kiro powers
├── .gemini/commands/ # Gemini CLI commands
├── AGENTS.md # Universal agent docs
-├── 00-inbox/ # Profiles & onboarding files
+├── CLAUDE.md # Framework instructions
+├── 00-inbox/ # Profiles, interests, integrations
├── 01-daily/ # Briefs & check-ins
├── 02-personal/ # Personal braindumps (private)
├── 03-professional/ # Professional braindumps & strategy
@@ -158,6 +164,7 @@ Git is optional but recommended for version history. COG works fine with just iC
- [x] ~~Gemini CLI + OpenAI Codex support~~ (shipped in v3.1)
- [x] ~~Upstream update system~~ (shipped in v3.2)
+- [x] ~~Role packs & integration discovery~~ (shipped in v3.3)
- [ ] Web interface for knowledge graph visualization
- [ ] Mobile-first commands (optimized for Obsidian mobile)
- [ ] Team collaboration features (with privacy preservation)
diff --git a/SETUP.md b/SETUP.md
index 993a0a6..68c6d95 100644
--- a/SETUP.md
+++ b/SETUP.md
@@ -42,7 +42,7 @@ That's it! You now have a working second brain.
**What just happened?**
- The cloned `COG-second-brain` folder IS your second brain
-- 6 AI skills are available in multiple formats:
+- 10 AI skills are available in multiple formats:
- `.claude/skills/` - For Claude Code
- `.kiro/powers/` - For Kiro
- `agents.md` - For any agent
@@ -81,11 +81,14 @@ Onboarding will ask you:
**Takes 2 minutes. Everything is stored as markdown files you can edit.**
Onboarding creates:
-- `00-inbox/MY-PROFILE.md` - Your info and projects
+- `00-inbox/MY-PROFILE.md` - Your info, role pack, and projects
- `00-inbox/MY-INTERESTS.md` - Topics for daily briefs
+- `00-inbox/MY-INTEGRATIONS.md` - Active/disabled external service integrations
- `03-professional/COMPETITIVE-WATCHLIST.md` - Tracking list (if any)
- Project folders in `04-projects/`
+**Role-based personalization:** COG matches your role to a role pack that prioritizes the most relevant skills and integrations for you. Available packs: Product Manager, Engineering Lead, Engineer, Designer, Founder, Marketer.
+
### Test Your Setup
**Try your first braindump:**
@@ -123,7 +126,8 @@ After running onboarding, you'll have this structure:
COG-second-brain/ # This is your second brain folder
├── agents.md # Universal agent documentation
├── .claude/
-│ └── skills/ # 6 Claude Code skills
+│ ├── roles/ # 7 role packs for personalized recommendations
+│ └── skills/ # 10 Claude Code skills
│ ├── onboarding/
│ ├── braindump/
│ ├── daily-brief/
@@ -138,7 +142,8 @@ COG-second-brain/ # This is your second brain folder
│ ├── cog-weekly-checkin/
│ ├── cog-knowledge-consolidation/
│ └── cog-url-dump/
-├── 00-inbox/ # Temporary & profile files (created by onboarding)
+├── CLAUDE.md # Framework instructions (role packs, integrations)
+├── 00-inbox/ # Profiles, interests, integrations (created by onboarding)
├── 01-daily/ # Daily briefs and check-ins
│ ├── briefs/
│ └── checkins/
diff --git a/cog-update.sh b/cog-update.sh
index e818268..4a2e4e2 100755
--- a/cog-update.sh
+++ b/cog-update.sh
@@ -38,6 +38,21 @@ FRAMEWORK_FILES=(
".claude/skills/knowledge-consolidation/SKILL.md"
".claude/skills/url-dump/SKILL.md"
".claude/skills/update-cog/SKILL.md"
+ ".claude/skills/team-brief/SKILL.md"
+ ".claude/skills/comprehensive-analysis/SKILL.md"
+ ".claude/skills/meeting-transcript/SKILL.md"
+
+ # Role packs
+ ".claude/roles/_template.md"
+ ".claude/roles/product-manager.md"
+ ".claude/roles/engineer.md"
+ ".claude/roles/engineering-lead.md"
+ ".claude/roles/designer.md"
+ ".claude/roles/founder.md"
+ ".claude/roles/marketer.md"
+
+ # Framework config
+ "CLAUDE.md"
# Kiro powers
".kiro/powers/cog-onboarding/POWER.md"
diff --git a/marketplace-entry.json b/marketplace-entry.json
index 0ddc5fc..8510a1c 100644
--- a/marketplace-entry.json
+++ b/marketplace-entry.json
@@ -7,5 +7,5 @@
"url": "https://github.com/huytieu"
},
"description": "Self-evolving second brain with AI agents for knowledge management",
- "version": "3.2.0"
+ "version": "3.3.0"
}