> ## Documentation Index
> Fetch the complete documentation index at: https://clawhqplatform.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent Packs

> Pre-built agent bundles for common workflows. Install in one click, customize with YAML.

Packs are pre-configured collections of agents, pipelines, and triggers. Instead of building from scratch, install a pack and your agents are ready to work.

## Pricing

| Tier        | Price | What's included                                        |
| ----------- | ----- | ------------------------------------------------------ |
| **Free**    | \$0   | Community packs — install and use forever              |
| **Starter** | \$29  | Single lightweight decision-making pack                |
| **Pro**     | \$49  | Lighter workflow packs for individuals and small teams |
| **Pro+**    | \$99  | Heavy multi-agent packs with advanced orchestration    |

***

## Available packs

### Coordination packs

| Pack                                 | Agents                                       | Best for                                                                       |
| ------------------------------------ | -------------------------------------------- | ------------------------------------------------------------------------------ |
| [Observer Pack](/guides/agent-teams) | Observer, Conflict Detector, Status Reporter | Teams running 3+ parallel agents who need coordination without an orchestrator |

### Team use case packs

Buy links are available at [clawhqplatform.com/packs](https://clawhqplatform.com/packs).

***

## Free packs

### Data Analyst

**Free**

Turn any CSV, JSON, spreadsheet, or database query into charts, tables, and statistical summaries — no code required.

**What it does:**

* Accepts raw data via file upload, paste, or tool output
* Generates bar, line, scatter, histogram, and pie charts
* Produces formatted tables with sorting and filtering
* Computes statistical summaries (mean, median, std dev, outliers, correlation)
* Outputs shareable reports to Slack, Discord, or Notion

**Included agents:** Data Reader, Chart Generator, Stats Analyst, Report Writer

```bash theme={null}
cp packs/data-analyst.yaml ~/.clawhq/packs/
docker compose restart paperclip
```

***

### Weekly Briefing

**Free**

Every Friday morning, synthesize everything across Slack, GitHub, Linear, and email into a crisp executive summary.

**What it does:**

* Pulls activity from all connected channels
* Identifies key decisions, blockers, and upcoming deadlines
* Surfaces threads that went quiet and may need a nudge
* Posts a structured digest to your chosen channel

**Schedule:** Runs every Friday at 8am by default. Configurable via cron.

**Included agents:** Digest Collector, Priority Ranker, Briefing Writer

***

### Content Engine

**Free**

Draft LinkedIn posts, tweets, newsletter sections, and blog outlines from your raw notes. Repurpose content across formats and maintain a consistent brand voice.

**What it does:**

* Drafts content from bullet points, voice notes, or rough text
* Adapts tone per platform
* Repurposes a single piece into 4–6 formats
* Maintains a brand voice profile across all output

**Included agents:** Content Drafter, Brand Voice Enforcer, Platform Adapter, Scheduler

***

### Innovation Scout

**Free**

Discover automation opportunities by merging your team's internal pain signals with external trends. Delivers a ranked opportunity report weekly.

**Included agents:** Pain Signal Collector, Trend Researcher, Opportunity Ranker

***

### Workflow Docs

**Free**

Automatically generate Tango-style step-by-step workflow documentation as agents work. Every multi-step task becomes a numbered SOP saved to your docs folder — no manual writeup required.

**What it does:**

* Detects when a task took 3+ distinct steps and auto-generates a guide
* Produces numbered steps with clear descriptions, expected outputs, and tips
* Supports live-mode recording and SOP-from-description generation
* Maintains a searchable index of all saved guides
* Saves to `~/docs/workflows/<slug>.md` — git-friendly markdown

**Activation phrases:** "document this", "write a guide", "create an SOP", "record what you're doing"

**Included agents:** workflow-docs

```bash theme={null}
cp packs/workflow-docs.yaml ~/.clawhq/packs/
docker compose restart paperclip
```

***

### LLM Wiki (Karpathy Knowledge)

**Free**

Build a persistent, compounding personal knowledge base as interlinked markdown wiki pages. Drop a URL, paste text, or share a file — the agent ingests it, writes structured pages, and maintains cross-references automatically.

**What it does:**

* Auto-initializes a wiki from the first source you give it — no setup required
* Ingests URLs, files, and pasted text into a structured `~/wiki/<topic>/` directory
* Creates entity, concept, and source pages with wikilinks and citations
* Detects contradictions between sources and flags them inline
* Runs lint passes to find orphan pages, missing links, and stale overviews
* Answers questions from the compiled knowledge with cited responses

**Activation phrases:** "start a wiki", "add this to the wiki", "ingest this", "what does my wiki say about X"

**Inspired by:** [Andrej Karpathy's LLM wiki pattern](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f)

**Included agents:** llm-wiki

```bash theme={null}
cp packs/llm-wiki.yaml ~/.clawhq/packs/
docker compose restart paperclip
```

***

## Starter packs (\$29)

### Decision Brief

**\$29**

Transform scattered, unstructured notes into one execution-ready decision brief with clear next actions — in under 2 minutes. Accepts raw notes, transcripts, Slack threads, interview notes, or any mixed-format input.

**What it does:**

* Normalizes input from any source format (notes, transcripts, chat logs, docs)
* Extracts signals: blockers, next actions, owners, and confidence scores
* Identifies evidence gaps and contradictions before recommending
* Applies weighted scoring: Urgency (40%), Impact (35%), Confidence (25%)
* Outputs a human-readable brief and a machine-ready JSON payload

**Output format:**

```
## Decision Brief — [topic]

### Key Findings
### Recommendation (with confidence %)
### Next Actions (owner + due date)
### Evidence Gaps
### Risk Flags
```

**Included agents:** decision-brief

```bash theme={null}
cp packs/decision-brief.yaml ~/.clawhq/packs/
docker compose restart paperclip
```

***

## Pro packs (\$49)

### Creator Suite

**\$49**

Full content production system — script writing, video outlining, thumbnail copy, SEO optimization, and multi-platform repurposing for creators and media teams.

**Included agents:** script-writer, video-outliner, thumbnail-copy, seo-optimizer, content-repurposer

**Works well with:** [Google Vids](https://vids.google.com) — Google's AI video tool now includes free VO3.1 video generation (10 clips/month with any Gmail), directable AI avatars, LIA 3 custom music generation (AI Pro/Ultra), a Chrome screen recorder, and direct YouTube publishing. Pair it with Creator Suite agents to go from script to published video inside one workflow.

***

### Reskilling Core

**\$49**

Personalized learning paths, skills gap analysis, and professional development planning. Helps individuals and teams identify what to learn and how to prioritize it.

**Included agents:** skills-assessor, learning-path-builder, progress-tracker, resource-curator

***

### Marketing Agency

**\$49**

Campaign planning, copy generation, asset briefing, and performance analysis for marketing teams and agencies managing multiple clients or campaigns.

**Included agents:** campaign-planner, copy-generator, brief-writer, performance-analyst

***

### Product Research

**\$49**

Full user research lifecycle — interview scripts, surveys, usability tests, personas, JTBD analysis, and findings synthesis. For product teams making data-informed decisions.

**Included agents:** research-planner, interview-scripter, survey-builder, findings-synthesizer, persona-builder, jtbd-analyst

***

## Pro+ packs (\$99)

### Finance Dashboards

**\$99**

Automated financial reporting, variance analysis, cash flow modeling, and board-ready dashboard generation from your accounting data.

**Included agents:** report-builder, variance-analyst, cashflow-modeler, dashboard-generator

***

### Executive Assistant

**\$99**

A full AI chief of staff. Manages your calendar, triages your inbox, coordinates travel, preps you for every meeting, and handles routine correspondence.

| Task                  | Detail                                                              |
| --------------------- | ------------------------------------------------------------------- |
| Calendar management   | Schedule meetings, resolve conflicts, block focus time              |
| Email triage          | Categorize by urgency, draft replies, flag action items             |
| Meeting prep          | Pull briefings, past notes, and open action items before every call |
| Travel coordination   | Book flights, hotels, ground transport — full itinerary in Slack    |
| Stakeholder briefings | Research contacts before meetings: role, history, key interests     |
| Task prioritization   | Morning briefing ranked by impact                                   |
| Expense tracking      | Log from receipts, categorize, generate monthly report              |

**Required MCPs:** Google Calendar or Outlook, Gmail or Outlook Mail, Notion (optional)

**Included agents:** Scheduler, Inbox Triage, Meeting Prep, Travel Coordinator, Briefing Writer

***

### Competitor Intel

**\$99**

Monitors competitor websites, pricing pages, job postings, GitHub activity, and social channels. Weekly digest lands in Slack with diffs from the prior week.

**What it does:**

* Watches for pricing, feature, and messaging changes
* Monitors job postings to infer product direction
* Tracks GitHub repos for new releases and capability shifts
* Delivers a structured weekly intel report

**Included agents:** Web Monitor, Job Board Watcher, GitHub Tracker, Intel Synthesizer

***

### Deal Room

**\$99**

Research prospects, draft personalized outreach sequences, prepare pre-call briefings, and log everything to your CRM — all triggered from a Slack message.

**Required MCPs:** LinkedIn (via Playwright), Salesforce or HubSpot, email

**Included agents:** Prospect Researcher, Outreach Drafter, Call Prep, CRM Logger

***

### Augmented Engineering

**\$99**

PR reviews, automated release notes, issue triage, test coverage analysis, and code quality reporting for engineering teams on GitHub or GitLab.

**Included agents:** pr-reviewer, release-notes-writer, issue-triager, coverage-analyst

***

### Augmented Finance

**\$99**

Invoice processing, expense auditing, reconciliation, and board report generation for finance and accounting teams.

**Included agents:** invoice-processor, expense-auditor, reconciler, board-report-generator

***

### Augmented Sales

**\$99**

Lead qualification, outreach sequencing, objection handling prep, and CRM hygiene automation for sales teams.

**Included agents:** lead-qualifier, sequence-builder, objection-prepper, crm-hygienist

***

### Augmented Support

**\$99**

Ticket classification, knowledge base search, reply drafting, and escalation routing for customer support teams.

**Included agents:** ticket-classifier, kb-search, reply-drafter, escalation-router

***

### Real Estate

**\$99**

Property research, listing analysis, market comparables, and client report generation for real estate professionals.

**Included agents:** property-researcher, comps-analyst, listing-writer, client-report-generator

***

### E-commerce

**\$99**

Product description generation, inventory monitoring, review analysis, and promotional copy for e-commerce teams.

**Included agents:** product-writer, inventory-monitor, review-analyst, promo-copy-generator

***

## Bundles

Bundles combine related packs at a discount. A good starting point if you know your use case.

### All Packs Bundle — \$299

Every paid pack in one purchase. Save \$916 vs buying separately.

[Get all packs →](https://buy.stripe.com/bJe28s2IP6ZheG94Vm2Fa0b)

***

### Sales Bundle — \$149

**Saves \$98** · Augmented Sales + Augmented Support + Marketing Agency

Everything a revenue team needs to prospect, close, and retain customers.

[Get Sales Bundle →](https://buy.stripe.com/4gMcN6fvB2J17dH87y2Fa0p)

***

### Finance Bundle — \$149

**Saves \$148** · Finance Dashboards + Augmented Finance + Executive Assistant

Automate financial reporting, model complex scenarios, and streamline executive workflows.

[Get Finance Bundle →](https://buy.stripe.com/aFa7sMerx0ATdC573u2Fa0q)

***

### Engineering Bundle — \$149

**Saves \$98** · Augmented Engineering + Product Research + Competitor Intel

From research to shipped product — deep research, competitive intelligence, and engineering acceleration.

[Get Engineering Bundle →](https://buy.stripe.com/fZudRa5V1gzR41v4Vm2Fa0r)

***

### Starter Bundle — \$79

**Saves \$48** · Creator Suite + Reskilling Core + Decision Brief

The best starting point for individuals and small teams leveling up with AI.

[Get Starter Bundle →](https://buy.stripe.com/dRm5kEerxdnFbtX1Ja2Fa0s)

***

## Install a pack

```bash theme={null}
# Copy the pack YAML to your ClawHQ workspace
cp packs/weekly-briefing.yaml ~/.clawhq/packs/

# Restart Paperclip to load the new pack
docker compose restart paperclip
```

Or from the dashboard: **Packs → Install**.

## Build your own

Packs follow the `paperclip.ai/v1 AgentPack` schema:

```yaml theme={null}
apiVersion: paperclip.ai/v1
kind: AgentPack
metadata:
  name: my-pack
  displayName: "My Custom Pack"

agents:
  - id: my-agent
    name: "My Agent"
    trigger:
      - type: channel
        keywords: ["help me with"]
    task: |
      You are a helpful agent. When someone asks for help...
    tools:
      - web_search
      - doc_writer
```

See the [Pack Schema Reference](/reference/pack-schema) for all available fields.

## Pack security vetting

Every pack uploaded to the ClawHQ registry is automatically vetted before it is stored. The vetter rejects packs that fail any hard check.

### Checks run on all packs

| Check                     | Level    | What it catches                                                   |
| ------------------------- | -------- | ----------------------------------------------------------------- |
| Schema fields present     | **FAIL** | Missing `apiVersion`, `kind`, `name`, `version`, or `description` |
| No hardcoded secrets      | **FAIL** | API keys, tokens, PEM blocks, or `password:` fields in YAML text  |
| Dangerous tools need hitl | **FAIL** | `bash`, `exec`, `fs_delete`, `cron` declared without `hitl: true` |
| External URLs in prompts  | WARN     | HTTP/S URLs in task or instructions text                          |
| Semver version            | WARN     | Version field not in `1.0.0` format                               |
| Tags declared             | WARN     | No tags (affects discoverability and audit filtering)             |

### Additional checks for third-party packs

Packs submitted from outside the ClawHQ team are subject to stricter rules:

| Check              | Level    | Requirement                                                           |
| ------------------ | -------- | --------------------------------------------------------------------- |
| Publisher identity | **FAIL** | `metadata.publisher` must be present and cannot be `clawhq`           |
| Contact or repo    | **FAIL** | `metadata.contact` (email) or `metadata.repository` (URL) required    |
| External URLs      | **FAIL** | Upgraded from warning to hard fail — no external URLs in task prompts |
| Tags               | **FAIL** | Tags are mandatory, not optional                                      |

### Vet packs locally before uploading

```bash theme={null}
# Vet all packs in the packs/ directory
npx tsx scripts/vet-pack.ts packs/

# Vet a single pack
npx tsx scripts/vet-pack.ts packs/my-pack.yaml

# Vet an external/third-party pack with stricter rules
npx tsx scripts/vet-pack.ts my-vendor-pack.yaml --third-party
```

Exit code `0` means all packs pass (warnings are allowed). Exit code `1` means at least one FAIL — the pack will be rejected by the registry.