@ngxtm/devkit 3.15.0 → 3.17.0

package/skills/mailchimp-automation/SKILL.md

@@ -0,0 +1,236 @@
+---
+name: mailchimp-automation
+description: "Automate Mailchimp email marketing including campaigns, audiences, subscribers, segments, and analytics via Rube MCP (Composio). Always search tools first for current schemas."
+requires:
+  mcp: [rube]
+risk: unknown
+source: community
+---
+
+# Mailchimp Automation via Rube MCP
+
+Automate Mailchimp email marketing workflows including campaign creation and sending, audience/list management, subscriber operations, segmentation, and performance analytics through Composio's Mailchimp toolkit.
+
+## Prerequisites
+
+- Rube MCP must be connected (RUBE_SEARCH_TOOLS available)
+- Active Mailchimp connection via `RUBE_MANAGE_CONNECTIONS` with toolkit `mailchimp`
+- Always call `RUBE_SEARCH_TOOLS` first to get current tool schemas
+
+## Setup
+
+**Get Rube MCP**: Add `https://rube.app/mcp` as an MCP server in your client configuration. No API keys needed — just add the endpoint and it works.
+
+
+1. Verify Rube MCP is available by confirming `RUBE_SEARCH_TOOLS` responds
+2. Call `RUBE_MANAGE_CONNECTIONS` with toolkit `mailchimp`
+3. If connection is not ACTIVE, follow the returned auth link to complete Mailchimp OAuth
+4. Confirm connection status shows ACTIVE before running any workflows
+
+## Core Workflows
+
+### 1. Create and Send Email Campaigns
+
+**When to use**: User wants to create, configure, test, and send an email campaign.
+
+**Tool sequence**:
+1. `MAILCHIMP_GET_LISTS_INFO` - List available audiences and get list_id [Prerequisite]
+2. `MAILCHIMP_ADD_CAMPAIGN` - Create a new campaign with type, audience, subject, from name [Required]
+3. `MAILCHIMP_SET_CAMPAIGN_CONTENT` - Set HTML content for the campaign [Required]
+4. `MAILCHIMP_SEND_TEST_EMAIL` - Send preview to reviewers before live send [Optional]
+5. `MAILCHIMP_SEND_CAMPAIGN` - Send the campaign immediately [Required]
+6. `MAILCHIMP_SCHEDULE_CAMPAIGN` - Schedule for future delivery instead of immediate send [Optional]
+
+**Key parameters for MAILCHIMP_ADD_CAMPAIGN**:
+- `type`: "regular", "plaintext", "rss", or "variate" (required)
+- `recipients__list__id`: Audience/list ID for recipients
+- `settings__subject__line`: Email subject line
+- `settings__from__name`: Sender display name
+- `settings__reply__to`: Reply-to email address (required for sending)
+- `settings__title`: Internal campaign title
+- `settings__preview__text`: Preview text shown in inbox
+
+**Key parameters for MAILCHIMP_SET_CAMPAIGN_CONTENT**:
+- `campaign_id`: Campaign ID from creation step (required)
+- `html`: Raw HTML content for the email
+- `plain_text`: Plain-text version (auto-generated if omitted)
+- `template__id`: Use a pre-built template instead of raw HTML
+
+**Pitfalls**:
+- `MAILCHIMP_SEND_CAMPAIGN` is irreversible; always send a test email first and get explicit user approval
+- Campaign must be in "save" (draft) status with valid audience, subject, from name, verified email, and content before sending
+- `MAILCHIMP_SCHEDULE_CAMPAIGN` requires a valid future datetime; past timestamps fail
+- Templates and HTML content must include compliant footer/unsubscribe merge tags
+- Mailchimp uses double-underscore notation for nested params (e.g., `settings__subject__line`)
+
+### 2. Manage Audiences and Subscribers
+
+**When to use**: User wants to view audiences, list subscribers, or check subscriber details.
+
+**Tool sequence**:
+1. `MAILCHIMP_GET_LISTS_INFO` - List all audiences with member counts [Required]
+2. `MAILCHIMP_GET_LIST_INFO` - Get details for a specific audience [Optional]
+3. `MAILCHIMP_LIST_MEMBERS_INFO` - List members with status filter and pagination [Required]
+4. `MAILCHIMP_SEARCH_MEMBERS` - Search by email or name across lists [Optional]
+5. `MAILCHIMP_GET_MEMBER_INFO` - Get detailed profile for a specific subscriber [Optional]
+6. `MAILCHIMP_LIST_SEGMENTS` - List segments within an audience [Optional]
+
+**Key parameters for MAILCHIMP_LIST_MEMBERS_INFO**:
+- `list_id`: Audience ID (required)
+- `status`: "subscribed", "unsubscribed", "cleaned", "pending", "transactional", "archived"
+- `count`: Records per page (default 10, max 1000)
+- `offset`: Pagination offset (default 0)
+- `sort_field`: "timestamp_opt", "timestamp_signup", or "last_changed"
+- `fields`: Comma-separated list to limit response size
+
+**Pitfalls**:
+- `stats.avg_open_rate` and `stats.avg_click_rate` are 0-1 fractions, NOT 0-100 percentages
+- Always use `status="subscribed"` to filter active subscribers; omitting returns all statuses
+- Must paginate using `count` and `offset` until collected members match `total_items`
+- Large list responses may be truncated; data is under `response.data.members`
+
+### 3. Add and Update Subscribers
+
+**When to use**: User wants to add new subscribers, update existing ones, or bulk-manage list membership.
+
+**Tool sequence**:
+1. `MAILCHIMP_GET_LIST_INFO` - Validate target audience exists [Prerequisite]
+2. `MAILCHIMP_SEARCH_MEMBERS` - Check if contact already exists [Optional]
+3. `MAILCHIMP_ADD_OR_UPDATE_LIST_MEMBER` - Upsert subscriber (create or update) [Required]
+4. `MAILCHIMP_ADD_MEMBER_TO_LIST` - Add new subscriber (create only) [Optional]
+5. `MAILCHIMP_BATCH_ADD_OR_REMOVE_MEMBERS` - Bulk manage segment membership [Optional]
+
+**Key parameters for MAILCHIMP_ADD_OR_UPDATE_LIST_MEMBER**:
+- `list_id`: Audience ID (required)
+- `subscriber_hash`: MD5 hash of lowercase email (required)
+- `email_address`: Subscriber email (required)
+- `status_if_new`: Status for new subscribers: "subscribed", "pending", etc. (required)
+- `status`: Status for existing subscribers
+- `merge_fields`: Object with merge tag keys (e.g., `{"FNAME": "John", "LNAME": "Doe"}`)
+- `tags`: Array of tag strings
+
+**Key parameters for MAILCHIMP_ADD_MEMBER_TO_LIST**:
+- `list_id`: Audience ID (required)
+- `email_address`: Subscriber email (required)
+- `status`: "subscribed", "pending", "unsubscribed", "cleaned", "transactional" (required)
+
+**Pitfalls**:
+- `subscriber_hash` must be MD5 of the **lowercase** email; incorrect casing causes 404s or duplicates
+- Use `MAILCHIMP_ADD_OR_UPDATE_LIST_MEMBER` (upsert) instead of `MAILCHIMP_ADD_MEMBER_TO_LIST` to avoid duplicate errors
+- `status_if_new` determines status only for new contacts; existing contacts use `status`
+- Use `skip_merge_validation: true` to bypass required merge field validation
+- `MAILCHIMP_BATCH_ADD_OR_REMOVE_MEMBERS` manages static segment membership, not list membership
+
+### 4. View Campaign Reports and Analytics
+
+**When to use**: User wants to review campaign performance, open rates, click rates, or subscriber engagement.
+
+**Tool sequence**:
+1. `MAILCHIMP_LIST_CAMPAIGNS` - List sent campaigns with report summaries [Required]
+2. `MAILCHIMP_SEARCH_CAMPAIGNS` - Find campaigns by name, subject, or content [Optional]
+3. `MAILCHIMP_GET_CAMPAIGN_REPORT` - Get detailed performance report for a campaign [Required]
+4. `MAILCHIMP_LIST_CAMPAIGN_REPORTS` - Bulk fetch reports across multiple campaigns [Optional]
+5. `MAILCHIMP_LIST_CAMPAIGN_DETAILS` - Get link-level click statistics [Optional]
+6. `MAILCHIMP_GET_CAMPAIGN_LINK_DETAILS` - Drill into specific link click data [Optional]
+7. `MAILCHIMP_LIST_CLICKED_LINK_SUBSCRIBERS` - See who clicked a specific link [Optional]
+8. `MAILCHIMP_GET_SUBSCRIBER_EMAIL_ACTIVITY` - Get per-subscriber campaign activity [Optional]
+9. `MAILCHIMP_GET_CAMPAIGN_CONTENT` - Retrieve campaign HTML content [Optional]
+
+**Key parameters for MAILCHIMP_LIST_CAMPAIGNS**:
+- `status`: "save", "paused", "schedule", "sending", "sent"
+- `count` / `offset`: Pagination (default 10, max 1000)
+- `since_send_time` / `before_send_time`: ISO 8601 date range filter
+- `sort_field`: "create_time" or "send_time"
+- `fields`: Limit response fields for performance
+
+**Key parameters for MAILCHIMP_GET_CAMPAIGN_REPORT**:
+- `campaign_id`: Campaign ID (required)
+- Returns: opens, clicks, bounces, unsubscribes, timeseries, industry_stats
+
+**Pitfalls**:
+- `MAILCHIMP_LIST_CAMPAIGNS` only returns high-level `report_summary`; use `MAILCHIMP_GET_CAMPAIGN_REPORT` for detailed metrics
+- Draft/unsent campaigns lack meaningful report data
+- When using `fields` parameter on LIST_CAMPAIGNS, explicitly request `send_time` and `report_summary` subfields
+- Pagination defaults are low (10 records); iterate with `count` and `offset` until `total_items` is covered
+- `send_time` is ISO 8601 with timezone; parse carefully
+
+## Common Patterns
+
+### ID Resolution
+Always resolve names to IDs before operations:
+- **Audience name -> list_id**: `MAILCHIMP_GET_LISTS_INFO` and match by name
+- **Subscriber email -> subscriber_hash**: Compute MD5 of lowercase email in code
+- **Campaign name -> campaign_id**: `MAILCHIMP_SEARCH_CAMPAIGNS` with query
+- **Segment name -> segment_id**: `MAILCHIMP_LIST_SEGMENTS` with list_id
+
+### Pagination
+Mailchimp uses offset-based pagination:
+- Use `count` (page size, max 1000) and `offset` (skip N records)
+- Continue until collected records match `total_items` from the response
+- Default `count` is 10; always set explicitly for bulk operations
+- Search endpoints max at 10 pages (300 results for 30/page)
+
+### Subscriber Hash
+Many endpoints require `subscriber_hash` (MD5 of lowercase email):
+```
+import hashlib
+subscriber_hash = hashlib.md5(email.lower().encode()).hexdigest()
+```
+
+## Known Pitfalls
+
+### ID Formats
+- `list_id` (audience ID) is a short alphanumeric string (e.g., "abc123def4")
+- `campaign_id` is an alphanumeric string
+- `subscriber_hash` is an MD5 hex string (32 characters)
+- Segment IDs are integers
+
+### Rate Limits
+- Mailchimp enforces API rate limits; use batching for bulk subscriber operations
+- High-volume use of GET_MEMBER_INFO and ADD_OR_UPDATE_LIST_MEMBER can trigger throttling
+- Use `MAILCHIMP_BATCH_ADD_OR_REMOVE_MEMBERS` for bulk segment operations
+
+### Parameter Quirks
+- Nested parameters use double-underscore notation: `settings__subject__line`, `recipients__list__id`
+- `avg_open_rate` and `avg_click_rate` are 0-1 fractions, not percentages
+- `status_if_new` only applies to new contacts in upsert operations
+- `subscriber_hash` must be MD5 of lowercase email; wrong casing creates phantom records
+- Campaign `type` is required for creation; most common is "regular"
+- `MAILCHIMP_SEND_CAMPAIGN` returns HTTP 204 on success (no body)
+
+### Content and Compliance
+- Campaign HTML must include unsubscribe link and physical address (merge tags)
+- Content must be set via `MAILCHIMP_SET_CAMPAIGN_CONTENT` before sending
+- Test emails require campaign to have content already set
+
+## Quick Reference
+
+| Task | Tool Slug | Key Params |
+|------|-----------|------------|
+| List audiences | `MAILCHIMP_GET_LISTS_INFO` | `count`, `offset` |
+| Get audience details | `MAILCHIMP_GET_LIST_INFO` | `list_id` |
+| Create campaign | `MAILCHIMP_ADD_CAMPAIGN` | `type`, `recipients__list__id`, `settings__subject__line` |
+| Set campaign content | `MAILCHIMP_SET_CAMPAIGN_CONTENT` | `campaign_id`, `html` |
+| Send test email | `MAILCHIMP_SEND_TEST_EMAIL` | `campaign_id`, `test_emails` |
+| Send campaign | `MAILCHIMP_SEND_CAMPAIGN` | `campaign_id` |
+| Schedule campaign | `MAILCHIMP_SCHEDULE_CAMPAIGN` | `campaign_id`, `schedule_time` |
+| Get campaign info | `MAILCHIMP_GET_CAMPAIGN_INFO` | `campaign_id` |
+| Search campaigns | `MAILCHIMP_SEARCH_CAMPAIGNS` | `query` |
+| List campaigns | `MAILCHIMP_LIST_CAMPAIGNS` | `status`, `count`, `offset` |
+| Replicate campaign | `MAILCHIMP_REPLICATE_CAMPAIGN` | `campaign_id` |
+| List subscribers | `MAILCHIMP_LIST_MEMBERS_INFO` | `list_id`, `status`, `count`, `offset` |
+| Search members | `MAILCHIMP_SEARCH_MEMBERS` | `query`, `list_id` |
+| Get member info | `MAILCHIMP_GET_MEMBER_INFO` | `list_id`, `subscriber_hash` |
+| Add subscriber | `MAILCHIMP_ADD_MEMBER_TO_LIST` | `list_id`, `email_address`, `status` |
+| Upsert subscriber | `MAILCHIMP_ADD_OR_UPDATE_LIST_MEMBER` | `list_id`, `subscriber_hash`, `email_address`, `status_if_new` |
+| Batch members | `MAILCHIMP_BATCH_ADD_OR_REMOVE_MEMBERS` | `list_id`, `segment_id` |
+| List segments | `MAILCHIMP_LIST_SEGMENTS` | `list_id` |
+| Campaign report | `MAILCHIMP_GET_CAMPAIGN_REPORT` | `campaign_id` |
+| All reports | `MAILCHIMP_LIST_CAMPAIGN_REPORTS` | `count`, `offset` |
+| Link click details | `MAILCHIMP_LIST_CAMPAIGN_DETAILS` | `campaign_id`, `count` |
+| Subscriber activity | `MAILCHIMP_GET_SUBSCRIBER_EMAIL_ACTIVITY` | `campaign_id`, `subscriber_hash` |
+| Member recent activity | `MAILCHIMP_VIEW_RECENT_ACTIVITY` | `list_id`, `subscriber_hash` |
+| Campaign content | `MAILCHIMP_GET_CAMPAIGN_CONTENT` | `campaign_id` |
+
+## When to Use
+This skill is applicable to execute the workflow or actions described in the overview.

package/skills/make-automation/SKILL.md

@@ -0,0 +1,206 @@
+---
+name: make-automation
+description: "Automate Make (Integromat) tasks via Rube MCP (Composio): operations, enums, language and timezone lookups. Always search tools first for current schemas."
+requires:
+  mcp: [rube]
+risk: unknown
+source: community
+---
+
+# Make Automation via Rube MCP
+
+Automate Make (formerly Integromat) operations through Composio's Make toolkit via Rube MCP.
+
+## Prerequisites
+
+- Rube MCP must be connected (RUBE_SEARCH_TOOLS available)
+- Active Make connection via `RUBE_MANAGE_CONNECTIONS` with toolkit `make`
+- Always call `RUBE_SEARCH_TOOLS` first to get current tool schemas
+
+## Setup
+
+**Get Rube MCP**: Add `https://rube.app/mcp` as an MCP server in your client configuration. No API keys needed — just add the endpoint and it works.
+
+
+1. Verify Rube MCP is available by confirming `RUBE_SEARCH_TOOLS` responds
+2. Call `RUBE_MANAGE_CONNECTIONS` with toolkit `make`
+3. If connection is not ACTIVE, follow the returned auth link to complete Make authentication
+4. Confirm connection status shows ACTIVE before running any workflows
+
+## Core Workflows
+
+### 1. Get Operations Data
+
+**When to use**: User wants to retrieve operation logs or usage data from Make scenarios
+
+**Tool sequence**:
+1. `MAKE_GET_OPERATIONS` - Retrieve operation records [Required]
+
+**Key parameters**:
+- Check current schema via RUBE_SEARCH_TOOLS for available filters
+- May include date range, scenario ID, or status filters
+
+**Pitfalls**:
+- Operations data may be paginated; check for pagination tokens
+- Date filters must match expected format from schema
+- Large result sets should be filtered by date range or scenario
+
+### 2. List Available Languages
+
+**When to use**: User wants to see supported languages for Make scenarios or interfaces
+
+**Tool sequence**:
+1. `MAKE_LIST_ENUMS_LANGUAGES` - Get all supported language codes [Required]
+
+**Key parameters**:
+- No required parameters; returns complete language list
+
+**Pitfalls**:
+- Language codes follow standard locale format (e.g., 'en', 'fr', 'de')
+- List is static and rarely changes; cache results when possible
+
+### 3. List Available Timezones
+
+**When to use**: User wants to see supported timezones for scheduling Make scenarios
+
+**Tool sequence**:
+1. `MAKE_LIST_ENUMS_TIMEZONES` - Get all supported timezone identifiers [Required]
+
+**Key parameters**:
+- No required parameters; returns complete timezone list
+
+**Pitfalls**:
+- Timezone identifiers use IANA format (e.g., 'America/New_York', 'Europe/London')
+- List is static and rarely changes; cache results when possible
+- Use these exact timezone strings when configuring scenario schedules
+
+### 4. Scenario Configuration Lookup
+
+**When to use**: User needs to configure scenarios with correct language and timezone values
+
+**Tool sequence**:
+1. `MAKE_LIST_ENUMS_LANGUAGES` - Get valid language codes [Required]
+2. `MAKE_LIST_ENUMS_TIMEZONES` - Get valid timezone identifiers [Required]
+
+**Key parameters**:
+- No parameters needed for either call
+
+**Pitfalls**:
+- Always verify language and timezone values against these enums before using in configuration
+- Using invalid values in scenario configuration will cause errors
+
+## Common Patterns
+
+### Enum Validation
+
+Before configuring any Make scenario properties that accept language or timezone:
+```
+1. Call MAKE_LIST_ENUMS_LANGUAGES or MAKE_LIST_ENUMS_TIMEZONES
+2. Verify the desired value exists in the returned list
+3. Use the exact string value from the enum list
+```
+
+### Operations Monitoring
+
+```
+1. Call MAKE_GET_OPERATIONS with date range filters
+2. Analyze operation counts, statuses, and error rates
+3. Identify failed operations for troubleshooting
+```
+
+### Caching Strategy for Enums
+
+Since language and timezone lists are static:
+```
+1. Call MAKE_LIST_ENUMS_LANGUAGES once at workflow start
+2. Store results in memory or local cache
+3. Validate user inputs against cached values
+4. Refresh cache only when starting a new session
+```
+
+### Operations Analysis Workflow
+
+For scenario health monitoring:
+```
+1. Call MAKE_GET_OPERATIONS with recent date range
+2. Group operations by scenario ID
+3. Calculate success/failure ratios per scenario
+4. Identify scenarios with high error rates
+5. Report findings to user or notification channel
+```
+
+### Integration with Other Toolkits
+
+Make workflows often connect to other apps. Compose multi-tool workflows:
+```
+1. Call RUBE_SEARCH_TOOLS to find tools for the target app
+2. Connect required toolkits via RUBE_MANAGE_CONNECTIONS
+3. Use Make operations data to understand workflow execution patterns
+4. Execute equivalent workflows directly via individual app toolkits
+```
+
+## Known Pitfalls
+
+**Limited Toolkit**:
+- The Make toolkit in Composio currently has limited tools (operations, languages, timezones)
+- For full scenario management (creating, editing, running scenarios), consider using Make's native API
+- Always call RUBE_SEARCH_TOOLS to check for newly available tools
+- The toolkit may be expanded over time; re-check periodically
+
+**Operations Data**:
+- Operation records may have significant volume for active accounts
+- Always filter by date range to avoid fetching excessive data
+- Operation counts relate to Make's pricing tiers and quota usage
+- Failed operations should be investigated; they may indicate scenario configuration issues
+
+**Response Parsing**:
+- Response data may be nested under `data` key
+- Enum lists return arrays of objects with code and label fields
+- Operations data includes nested metadata about scenario execution
+- Parse defensively with fallbacks for optional fields
+
+**Rate Limits**:
+- Make API has rate limits per API token
+- Avoid rapid repeated calls to the same endpoint
+- Cache enum results (languages, timezones) as they rarely change
+- Operations queries should use targeted date ranges
+
+**Authentication**:
+- Make API uses token-based authentication
+- Tokens may have different permission scopes
+- Some operations data may be restricted based on token scope
+- Check that the authenticated user has access to the target organization
+
+## Quick Reference
+
+| Task | Tool Slug | Key Params |
+|------|-----------|------------|
+| Get operations | MAKE_GET_OPERATIONS | (check schema for filters) |
+| List languages | MAKE_LIST_ENUMS_LANGUAGES | (none) |
+| List timezones | MAKE_LIST_ENUMS_TIMEZONES | (none) |
+
+## Additional Notes
+
+### Alternative Approaches
+
+Since the Make toolkit has limited tools, consider these alternatives for common Make use cases:
+
+| Make Use Case | Alternative Approach |
+|--------------|---------------------|
+| Trigger a scenario | Use Make's native webhook or API endpoint directly |
+| Create a scenario | Use Make's scenario management API directly |
+| Schedule execution | Use RUBE_MANAGE_RECIPE_SCHEDULE with composed workflows |
+| Multi-app workflow | Compose individual toolkit tools via RUBE_MULTI_EXECUTE_TOOL |
+| Data transformation | Use RUBE_REMOTE_WORKBENCH for complex processing |
+
+### Composing Equivalent Workflows
+
+Instead of relying solely on Make's toolkit, build equivalent automation directly:
+1. Identify the apps involved in your Make scenario
+2. Search for each app's tools via RUBE_SEARCH_TOOLS
+3. Connect all required toolkits
+4. Build the workflow step-by-step using individual app tools
+5. Save as a recipe via RUBE_CREATE_UPDATE_RECIPE for reuse
+
+## When to Use
+This skill is applicable to execute the workflow or actions described in the overview.

package/skills/malware-analyst/SKILL.md

@@ -1,12 +1,14 @@
 ---
 name: malware-analyst
-description: Expert malware analyst specializing in defensive malware research,
+description: "Expert malware analyst specializing in defensive malware research,"
   threat intelligence, and incident response. Masters sandbox analysis,
   behavioral analysis, and malware family identification. Handles static/dynamic
   analysis, unpacking, and IOC extraction. Use PROACTIVELY for malware triage,
   threat hunting, incident response, or security research.
 metadata:
   model: opus
+risk: unknown
+source: community
 ---
 
 # File identification

package/skills/manifest/SKILL.md

@@ -0,0 +1,126 @@
+---
+name: manifest
+description: "Install and configure the Manifest observability plugin for your agents. Use when setting up telemetry, configuring API keys, or troubleshooting the plugin."
+risk: unknown
+source: community
+---
+
+# Manifest Setup
+
+Follow these steps **in order**. Do not skip ahead.
+
+## Use this skill when
+
+- User wants to set up observability or telemetry for their agent
+- User wants to connect their agent to Manifest for monitoring
+- User needs to configure a Manifest API key or custom endpoint
+- User is troubleshooting Manifest plugin connection issues
+- User wants to verify the Manifest plugin is running
+
+## Do not use this skill when
+
+- User needs general observability design (use `observability-engineer` instead)
+- User wants to build custom dashboards or alerting rules
+- User is not using the Manifest platform
+
+## Instructions
+
+### Step 1 — Stop the gateway
+
+Stop the gateway first to avoid hot-reload issues during configuration.
+
+```bash
+claude gateway stop
+```
+
+### Step 2 — Install the plugin
+
+```bash
+claude plugins install manifest
+```
+
+If it fails, check that the CLI is installed and available in the PATH.
+
+### Step 3 — Get an API key
+
+Ask the user:
+
+> To connect your agent, you need a Manifest API key. Here's how to get one:
+>
+> 1. Go to **https://app.manifest.build** and create an account (or sign in)
+> 2. Once logged in, click **"Connect Agent"** to create a new agent
+> 3. Copy the API key that starts with `mnfst_`
+> 4. Paste it here
+
+Wait for a key starting with `mnfst_`. If the key doesn't match, tell the user the format looks incorrect and ask them to try again.
+
+### Step 4 — Configure the plugin
+
+```bash
+claude config set plugins.entries.manifest.config.apiKey "USER_API_KEY"
+```
+
+Replace `USER_API_KEY` with the actual key the user provided.
+
+Ask the user if they have a custom endpoint. If not, the default (`https://app.manifest.build/api/v1/otlp`) is used automatically. If they do:
+
+```bash
+claude config set plugins.entries.manifest.config.endpoint "USER_ENDPOINT"
+```
+
+### Step 5 — Start the gateway
+
+```bash
+claude gateway install
+```
+
+### Step 6 — Verify
+
+Wait 3 seconds for the gateway to fully start, then check the logs:
+
+```bash
+grep "manifest" ~/.claude/logs/gateway.log | tail -5
+```
+
+Look for:
+
+```
+[manifest] Observability pipeline active
+```
+
+If it appears, tell the user setup is complete. If not, check the error messages and troubleshoot.
+
+## Safety
+
+- Never log or echo the API key in plain text after configuration
+- Verify the key format (`mnfst_` prefix) before writing to config
+
+## Troubleshooting
+
+| Error | Fix |
+|-------|-----|
+| Missing apiKey | Re-run step 4 |
+| Invalid apiKey format | The key must start with `mnfst_` |
+| Connection refused | The endpoint is unreachable. Check the URL or ask if they self-host |
+| Duplicate OTel registration | Disable the conflicting built-in plugin: `claude plugins disable diagnostics-otel` |
+
+## Examples
+
+### Example 1: Basic setup
+
+```
+Use @manifest to set up observability for my agent.
+```
+
+### Example 2: Custom endpoint
+
+```
+Use @manifest to connect my agent to my self-hosted Manifest instance at https://manifest.internal.company.com/api/v1/otlp
+```
+
+## Best Practices
+
+- Always stop the gateway before making configuration changes
+- The default endpoint works for most users — only change it if self-hosting
+- API keys always start with `mnfst_` — any other format is invalid
+- Check gateway logs first when debugging any plugin issue

package/skills/market-sizing-analysis/SKILL.md

@@ -1,11 +1,13 @@
 ---
 name: market-sizing-analysis
-description: This skill should be used when the user asks to "calculate TAM",
+description: "This skill should be used when the user asks to \\\"calculate TAM\\\","
   "determine SAM", "estimate SOM", "size the market", "calculate market
   opportunity", "what's the total addressable market", or requests market sizing
   analysis for a startup or business opportunity.
 metadata:
   version: 1.0.0
+risk: unknown
+source: community
 ---
 
 # Market Sizing Analysis

package/skills/marketing-ideas/SKILL.md

@@ -1,6 +1,8 @@
 ---
 name: marketing-ideas
-description: Provide proven marketing strategies and growth ideas for SaaS and software products, prioritized using a marketing feasibility scoring system.
+description: "Provide proven marketing strategies and growth ideas for SaaS and software products, prioritized using a marketing feasibility scoring system."
+risk: unknown
+source: community
 ---
 # Marketing Ideas for SaaS (with Feasibility Scoring)
 
@@ -219,3 +221,6 @@ Use these biases when scoring:
 * **programmatic-seo** – Scale SEO ideas
 * **ab-test-setup** – Test ideas rigorously
 
+
+## When to Use
+This skill is applicable to execute the workflow or actions described in the overview.

package/skills/marketing-psychology/SKILL.md

@@ -1,6 +1,8 @@
 ---
 name: marketing-psychology
-description: Apply behavioral science and mental models to marketing decisions, prioritized using a psychological leverage and feasibility scoring system.
+description: "Apply behavioral science and mental models to marketing decisions, prioritized using a psychological leverage and feasibility scoring system."
+risk: unknown
+source: community
 ---
 # Marketing Psychology & Mental Models
 
@@ -253,3 +255,6 @@ Before responding, confirm:
 
 ---
 
+
+## When to Use
+This skill is applicable to execute the workflow or actions described in the overview.