Multi-source AI profile intelligence — cross-reference your data, discover your hidden strengths.

---

Resume analyzers look at one source. Clawsight cross-references all of them.

Your resume says "Java lead" but your GitHub is 90% Go. Three LinkedIn recommenders mention mentoring, but your resume never says "leadership." You have payment systems + distributed architecture + global experience — a triple combination that's extremely rare. These insights are invisible from any single source. They only emerge when you look across data streams.

Named after the Mantis Shrimp (螳螂虾) — the creature with 16 types of color receptors that sees dimensions invisible to all other animals. Clawsight does the same for your professional identity.

Quick Start

# Install (one-time)
/install clawsight

# Import sources
/clawsight resume.pdf
/clawsight https://github.com/yourusername

# Discover insights (needs 2+ sources)
/clawsight insight

# Map your future potential
/clawsight potential

Demo Output

After importing a resume + GitHub profile, /clawsight insight produces:

╔══════════════════════════════════════════════════════════╗
║  🔍 Cross-Source Insight: Behavioral-Declarative Gap     ║
╠══════════════════════════════════════════════════════════╣
║  Resume: "Backend Engineer, Java"                        ║
║  GitHub: 47 repos, 92% Go, contributor to 3 CNCF projects║
║                                                          ║
║  → Your identity has shifted to Go/cloud-native, but     ║
║    your resume still tells a Java story.                 ║
║    This gap may cost you roles you're already qualified  ║
║    for.                                                  ║
╚══════════════════════════════════════════════════════════╝

╔══════════════════════════════════════════════════════════╗
║  💎 Compound Advantage (rare combination)                ║
╠══════════════════════════════════════════════════════════╣
║  Payment Systems × Distributed Arch × Global Ops         ║
║  Market rarity: top 2% of senior engineers               ║
║                                                          ║
║  → This triple stack is in extreme demand for fintech    ║
║    companies expanding internationally.                  ║
╚══════════════════════════════════════════════════════════╝

What It Does

Clawsight operates through three intelligence layers:

Layer	Command	What You Get
Profile	/clawsight <source>	Multi-source import with cross-correlation
Insight	/clawsight insight	Hidden strengths, blind spots, behavioral-declarative gaps
Potential	/clawsight potential	Compound advantages × industry trends mapping

Supported Sources

Source	Command	What It Captures
Resume (PDF/text)	/clawsight resume.pdf	Career narrative, skills, achievements
GitHub	/clawsight https://github.com/user	Real tech stack, contribution patterns, projects
LinkedIn export	/clawsight linkedin.zip	Endorsements, recommendations, network signals
Personal website	/clawsight https://yoursite.com	Self-presentation, projects, writing
JSON Resume	/clawsight resume.json	Structured profile data

LinkedIn Note: Automated access is blocked by LinkedIn. Export your data via Settings → Get a copy of your data — Clawsight will parse the ZIP. See docs/linkedin-guide.md for step-by-step instructions.

Additional Commands

Command	Description
/clawsight score	Profile completeness and understanding level
/clawsight refresh	Re-fetch all sources, detect changes over time

How It Works

 ┌─────────┐  ┌─────────┐  ┌─────────┐  ┌─────────┐
 │ Resume  │  │ GitHub  │  │LinkedIn │  │ Website │
 └────┬────┘  └────┬────┘  └────┬────┘  └────┬────┘
      │            │            │             │
      └──────────┬─┴────────────┴─────────────┘
                 ▼
       ┌─────────────────────┐
       │  Parse & Normalize  │
       └─────────┬───────────┘
                 ▼
       ┌─────────────────────┐
       │  Cross-Source        │  ★ Core differentiator
       │  Reconciliation      │  Classifies 5 conflict types,
       │  Engine              │  resolves 4 automatically,
       └─────────┬───────────┘  turns contradictions → insights
                 ▼
       ┌─────────────────────┐
       │  ⛔ Privacy Preview  │  User confirms before any write
       └─────────┬───────────┘
                 ▼
    ┌────────────┼────────────┐
    ▼            ▼            ▼
┌────────┐ ┌──────────┐ ┌───────────┐
│USER.md │ │MEMORY.md │ │projects/* │
└────────┘ └──────────┘ └───────────┘
                 │
        ┌────────┴────────┐
        ▼                 ▼
   ┌─────────┐     ┌───────────┐
   │ Insight │     │ Potential │
   │ Engine  │     │ Discovery │
   └─────────┘     └───────────┘

See docs/architecture.md for the full technical deep dive.

Scene Skills

Clawsight generates profile data. Scene Skills consume it for specialized analysis:

Skill	Description
career-mirror	Career direction analysis with compound advantage mapping

Planned Scene Skills

• tech-compass — Technology learning roadmap based on your stack + market trends
• writing-voice — Write content in your authentic voice
• learning-path — Personalized learning plan from skill gaps
• stakeholder-briefer — Quick context briefs for new collaborators

Documentation

Doc	Content
Architecture	System design, pipeline, data flow
Schema	Canonical extraction schema
Scoring	Profile completeness scoring
Templates	Output templates
User Journey	Interaction lifecycle
LinkedIn Guide	LinkedIn data export steps
Changelog	Version history

Roadmap

• v0.3 — Multi-source engine + Pure Skill rewrite
• v0.4 — Insight deepening + LinkedIn recommendations + refresh
• v0.5 — Potential discovery + dialogue enrichment + career-mirror
• v1.0 — OpenClaw profile standard + 5 Scene Skills + Hindsight integration

Contributing

Contributions are welcome! Here's how you can help:

1. Build a Scene Skill — Create a new skill under skills/ that consumes Clawsight's profile data for specialized analysis. See career-mirror as a reference.
2. Add a data source parser — Propose a new source type (e.g., Stack Overflow, Dribbble) by opening an issue.
3. Improve cross-source heuristics — The reconciliation rules in SKILL.md can always be sharpened with real-world edge cases.
4. Fix bugs & docs — Typos, unclear instructions, broken examples — all PRs welcome.

# Fork → Clone → Create branch
git checkout -b feature/my-scene-skill

# Make changes, then
git push origin feature/my-scene-skill
# Open a PR

License

MIT — use it, modify it, ship it.