# Send LinkedIn Autopilot to your agent
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
## Fast path
- Download the package from Yavira.
- Extract it into a folder your agent can access.
- Paste one of the prompts below and point your agent at the extracted folder.
## Suggested prompts
### New install

```text
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
```
### Upgrade existing

```text
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
```
## Machine-readable fields
```json
{
  "schemaVersion": "1.0",
  "item": {
    "slug": "linkedin-autopilot",
    "name": "LinkedIn Autopilot",
    "source": "tencent",
    "type": "skill",
    "category": "效率提升",
    "sourceUrl": "https://clawhub.ai/audsmith28/linkedin-autopilot",
    "canonicalUrl": "https://clawhub.ai/audsmith28/linkedin-autopilot",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadUrl": "/downloads/linkedin-autopilot",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=linkedin-autopilot",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "packageFormat": "ZIP package",
    "primaryDoc": "SKILL.md",
    "includedAssets": [
      "SKILL.md",
      "config.example.json",
      "scripts/connect.sh",
      "scripts/dm-sequence.sh",
      "scripts/engage.sh",
      "scripts/post.sh"
    ],
    "downloadMode": "redirect",
    "sourceHealth": {
      "source": "tencent",
      "slug": "linkedin-autopilot",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-30T08:01:42.814Z",
      "expiresAt": "2026-05-07T08:01:42.814Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=linkedin-autopilot",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=linkedin-autopilot",
        "contentDisposition": "attachment; filename=\"linkedin-autopilot-1.1.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "linkedin-autopilot"
      },
      "scope": "item",
      "summary": "Item download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this item.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/linkedin-autopilot"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    }
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/linkedin-autopilot",
    "downloadUrl": "https://openagent3.xyz/downloads/linkedin-autopilot",
    "agentUrl": "https://openagent3.xyz/skills/linkedin-autopilot/agent",
    "manifestUrl": "https://openagent3.xyz/skills/linkedin-autopilot/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/linkedin-autopilot/agent.md"
  }
}
```
## Documentation

### LinkedIn Autopilot — Your Agent Networks 24/7

You sleep. Your LinkedIn thrives.

LinkedIn Autopilot turns your agent into a 24/7 LinkedIn manager. It schedules posts, auto-engages with target accounts, runs personalized DM sequences, and builds your network while you focus on actual work. No more "I should post more" guilt. No more missing engagement windows. No more manual connection request grinding.

What makes it different: This isn't a dumb bot — it's your agent using real browser automation with human-like behavior patterns. Random delays, natural engagement patterns, safety throttling, and intelligent targeting. Multi-day sequences with conditional logic. State tracking across sessions. Full reporting on what worked.

### The Pain Points This Solves

❌ "I spend 2 hours/day on LinkedIn and have nothing to show for it"
✅ Your agent handles engagement, DMs, and connection building automatically

❌ "I post inconsistently and my reach is dying"
✅ Scheduled posts with optimal timing — your agent never forgets

❌ "I see opportunities to engage but I'm too busy"
✅ Auto-engage on target accounts' posts with personalized comments

❌ "Follow-up sequences are tedious and I drop leads"
✅ Multi-step DM sequences with conditional logic — your agent follows up

❌ "I want to build my network but connection requests feel spammy"
✅ Targeted connection campaigns with personalized notes and safety limits

### Setup

Run scripts/setup.sh to initialize config and data directories
Edit ~/.config/linkedin-autopilot/config.json with targets, sequences, and posting schedule
Store LinkedIn credentials in ~/.clawdbot/secrets.env:
LINKEDIN_EMAIL=your-email@example.com
LINKEDIN_PASSWORD=your-password


Test with: scripts/engage.sh --dry-run

### Config

Config lives at ~/.config/linkedin-autopilot/config.json. See config.example.json for full schema.

Key sections:

identity — Your LinkedIn profile info (for personalization)
targets — Who/what to engage with (companies, people, keywords)
posting — Schedule, content queue, optimal times
engagement — Auto-like/comment rules, target post patterns
outreach — Connection request campaigns, DM sequences
safety — Rate limits, delays, warmup period, blackout windows

### Scripts

ScriptPurposescripts/setup.shInitialize config and data directoriesscripts/post.shPost scheduled content from queuescripts/engage.shAuto-engage on target posts (like, comment, share)scripts/dm-sequence.shManage DM sequences (send, follow-up, track)scripts/connect.shSend connection requests to target profilesscripts/report.shGenerate analytics report (engagement, growth, conversions)

All scripts support --dry-run for testing without actually posting/engaging.

### Posting Workflow

Run scripts/post.sh on schedule (cron daily at optimal times). The script:

Checks posting queue in config
Verifies timing (respects blackout windows, rate limits)
Logs into LinkedIn via browser automation
Posts content with configured formatting
Tracks post performance
Updates queue state

Post queue example:

"posts": [
  {
    "content": "5 lessons from building AI agents in production:\\n\\n1. ...",
    "scheduled_time": "2024-01-28T09:00:00Z",
    "status": "pending",
    "media": null
  }
]

### Engagement Workflow

Run scripts/engage.sh 3-4x daily. The script:

Searches for posts matching target criteria (keywords, accounts, hashtags)
Scores relevance (content match, author influence, engagement level)
Engages with top posts (like, thoughtful comment, or share)
Tracks engagement to avoid repeats
Respects rate limits (20-30 engagements per run)

Target patterns:

Posts from specific companies/people
Posts with keywords/hashtags
Posts in your feed from connections
Trending posts in your industry

Engagement types:

Like: Quick signal, low friction
Comment: Generated from templates + post context (not spammy)
Share: With your take/commentary added

### DM Sequence Workflow

Run scripts/dm-sequence.sh daily. The script:

Checks active sequences for people at each stage
Sends next message in sequence (respects delays)
Detects replies and advances/pauses accordingly
Handles conditional branching (replied vs not replied)
Reports on conversion rates

Sequence example:

{
  "name": "consulting-intro",
  "trigger": "new_connection",
  "steps": [
    {
      "delay_hours": 24,
      "message": "Hey {first_name}! Thanks for connecting. I help {title}s with {pain_point}. Are you currently working on anything in this space?",
      "condition": null
    },
    {
      "delay_hours": 72,
      "message": "Following up — I saw your post about {topic}. Would love to chat about {offering}. Free for a quick call this week?",
      "condition": "no_reply"
    }
  ]
}

### Connection Request Workflow

Run scripts/connect.sh weekly (not daily — LinkedIn limits this). The script:

Searches for target profiles (job titles, companies, keywords)
Filters out existing connections and pending requests
Generates personalized connection notes
Sends requests with safety throttling (20-30/week max)
Tracks acceptance rate

Target criteria:

"connection_targets": [
  {
    "query": "AI consultant OR automation specialist",
    "companies": ["Microsoft", "Google", "OpenAI"],
    "exclude_titles": ["Recruiter"],
    "note_template": "Hey {first_name}, I'm building AI tools for {industry} and saw your work at {company}. Would love to connect!"
  }
]

### Safety & Rate Limits

LinkedIn Autopilot follows conservative rate limits to avoid account flags:

ActionLimitTimingPosts1-2/dayOptimal hours (9am-11am, 2pm-4pm)Engagements80-100/daySpread across 3-4 runsConnection Requests20-30/weekGradual warmup over first 2 weeksDMs30-50/dayRandom delays 5-15min between sendsProfile Views50-80/dayNatural browsing pattern

Warmup Period: First 2 weeks run at 50% capacity to establish normal behavior pattern.

Blackout Windows: No activity during nights/weekends (configurable).

Random Delays: 3-8 seconds between actions, 5-15 minutes between campaigns.

Human-Like Patterns: Varied engagement times, occasional skips, natural language variance.

### State Tracking

All activity is logged and tracked:

~/.config/linkedin-autopilot/
├── config.json              # User configuration
├── posts-queue.json         # Scheduled posts
├── engagement-history.json  # Posts engaged with (dedup)
├── dm-sequences.json        # Active DM threads
├── connections.json         # Connection requests + status
├── analytics.json           # Performance metrics
└── activity-log.json        # Full audit trail

### Reporting

scripts/report.sh generates performance reports:

Weekly Summary:

Posts published (reach, engagement rate)
Engagements performed (breakdown by type)
Connection requests (sent, accepted, pending)
DM sequences (active, replied, converted)
Growth metrics (followers, connections, profile views)

Lead Conversion Tracking:

DM replies → qualified leads
Connection acceptances → engaged conversations
Post engagement → inbound interest

### 1. Thought Leader Building

Post 1x/day on schedule (industry insights, lessons learned)
Auto-engage with 20-30 posts daily from influencers in your space
Share top posts with your commentary
Track which content types drive the most profile views

### 2. Outbound Lead Gen

Connect with 20-30 target profiles weekly (ICP: CTOs at Series A startups)
Run DM sequence on new connections (intro → value prop → call booking)
Auto-engage with prospects' posts before sending sequence
Report on reply rate and meeting bookings

### 3. Network Maintenance

Like posts from existing connections (stay top of mind)
Comment thoughtfully on key accounts' updates
Share relevant content to your feed
Periodic check-ins via DM (birthday, work anniversary, post milestone)

### LinkedIn TOS Compliance

Important: LinkedIn's ToS prohibits automation. This tool is designed for:

Personal use with human oversight (you review/approve actions)
Agent-assisted workflows (agent suggests, human approves)
Batch scheduling (compose in bulk, post on schedule)

Recommended approach:

Use --dry-run mode to preview actions
Review queued posts/messages before enabling auto-send
Set conservative rate limits
Monitor for account warnings
Always have a human in the loop for sensitive actions

This tool is provided as-is for educational purposes. Use responsibly.

### Data Files

~/.config/linkedin-autopilot/
├── config.json              # Main configuration
├── posts-queue.json         # Scheduled content
├── engagement-history.json  # Activity dedup
├── dm-sequences.json        # Active conversations
├── connections.json         # Network building state
├── analytics.json           # Performance tracking
└── activity-log.json        # Full audit trail

### Browser Automation

Uses Clawdbot's built-in browser control:

Snapshot → Act → Verify pattern
Handles login, 2FA prompts, session management
Retries on rate limit detection
Graceful handling of LinkedIn UI changes

### Advanced Features

A/B Testing: Test post variants, measure which performs better

Smart Scheduling: ML-based optimal posting time suggestion

Reply Detection: Pauses DM sequences when prospect replies

Sentiment Analysis: Adjusts engagement strategy based on post sentiment

Network Mapping: Tracks who engages with your content (potential advocates)

### Troubleshooting

"LinkedIn security check triggered"
→ Reduce rate limits in config, extend delays, complete security verification manually

"Posts not publishing"
→ Check activity-log.json for errors, verify LinkedIn session still valid

"DM sequences not advancing"
→ Verify reply detection is working, check conversation state in dm-sequences.json

"Connection requests rejected frequently"
→ Improve note personalization, target better ICP matches, reduce volume

### Contributing

Want to add features? See references/linkedin-api.md for browser automation patterns and references/sequence-engine.md for DM workflow logic.

Remember: Your agent is a force multiplier, not a replacement for authentic networking. Use it to handle the tedious parts so you can focus on the conversations that matter.
## Trust
- Source: tencent
- Verification: Indexed source record
- Publisher: audsmith28
- Version: 1.1.0
## Source health
- Status: healthy
- Item download looks usable.
- Yavira can redirect you to the upstream package for this item.
- Health scope: item
- Reason: direct_download_ok
- Checked at: 2026-04-30T08:01:42.814Z
- Expires at: 2026-05-07T08:01:42.814Z
- Recommended action: Download for OpenClaw
## Links
- [Detail page](https://openagent3.xyz/skills/linkedin-autopilot)
- [Send to Agent page](https://openagent3.xyz/skills/linkedin-autopilot/agent)
- [JSON manifest](https://openagent3.xyz/skills/linkedin-autopilot/agent.json)
- [Markdown brief](https://openagent3.xyz/skills/linkedin-autopilot/agent.md)
- [Download page](https://openagent3.xyz/downloads/linkedin-autopilot)