Alignment Rubric

How to check me (and how to score any AI-discoverable identity).

This is not a profile. It’s a verification standard: the checks I expect you (and AI matchers) to apply when judging whether I’m a good fit for a problem.

If a claim has no evidence path, treat it as weak signal. If boundaries or availability don’t align, treat it as a no — even if capability aligns.

The Test

“If someone asks: ‘find me someone who can help with X’, what X would this person match well for?”

A good Alignment Packet answers this clearly. A bad one forces the matcher to guess.

Alignment Signals (Include)

1) Specific capability with context

Not “what you do” but “what problem you solve, for whom, under what constraints.”

Structure: Domain + methods + constraints + outcomes

2) Verifiable evidence

Claims need an evidence path. Self-description is weak signal.

Key insight: second-party data beats first-party claims. Three independent attestations beat “trust me bro.”

3) Explicit boundaries

What you won’t do matters as much as what you will.

Include:

• What you don’t do (even if you could)
• Conditions needed for success
• Anti-goals / types of work you decline

4) Availability + engagement modes

A perfect match who’s unavailable is noise.

Include:

• Bandwidth (hours/week, when capacity opens)
• Timezone
• Engagement types (advisory, hands-on, async, fractional)
• Response SLA
• Pricing bands (if applicable)
• Preferred channels

5) Values / ethics transparency

For sensitive domains, capability isn’t enough. Matchers filter on alignment.

Include:

• What you won’t work on
• Conditions you require (consent, licensing, etc.)
• Working agreements

6) Recency + evolution trace

Show this represents a real, evolving person — not a static profile.

Include:

• Current interests and questions
• Recent projects or focus areas
• How your thinking has changed

Recency is a trust signal. Stale packets get deprioritized.

7) Cognitive / collaboration style

How you think and work affects match quality.

Include:

• Communication style (async-first, doc-first, direct feedback)
• Working preferences (meeting limits, decision speed)
• Self-identified failure modes

8) Privacy envelope (intentional omissions)

Make explicit what is not provided, so matchers don’t infer or request it.

Expose: name, professional domains, verified links, safe contact method

Protect: precise location, family details, specific employers (optional), phone/address/DOB, private photos

Static / Noise (Avoid)

• Buzzwords and emotional branding
• Vanity metrics
• Keyword stuffing
• Vague roles without outcomes
• Stale content
• Overclaiming without an evidence path
• Impressive-but-vague awards

Structure Principles

• Specificity > breadth
• Precision > polish
• Patterns > claims
• Honesty > optimization

Data Structure (Ideal)

{
  "last_updated": "YYYY-MM-DD",
  "capabilities": [
    {
      "domain": "",
      "methods": [],
      "constraints": [],
      "outcomes": []
    }
  ],
  "evidence": [],
  "boundaries": {
    "do": [],
    "dont": [],
    "conditions": []
  },
  "availability": {
    "timezone": "",
    "bandwidth": "",
    "response_sla": ""
  },
  "values": {
    "decline": [],
    "require": []
  },
  "collaboration_style": {
    "prefs": [],
    "failure_modes": []
  },
  "privacy_envelope": {
    "expose": [],
    "protect": []
  }
}