AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

interpreting-culture-index

Interprets Culture Index (CI) surveys, behavioral profiles, and personality assessment data. Supports individual profile interpretation, team composition analysis (gas/brake/glue), burnout detection, profile comparison, hiring profiles, manager coaching, interview transcript analysis for trait prediction, candidate debrief, onboarding planning, and conflict mediation. Accepts extracted JSON or PDF input via OpenCV extraction script. Use when the user shares a Culture Index PDF or JSON profile, asks what someone's CI traits mean, compares Culture Index profiles across a team or against a hiring profile, or asks about burnout risk from Survey-versus-Job gaps.

Original source
trailofbits
License
CC-BY-SA-4.0
Pinned revision
2be6433b39e009cd2b6fd3be1895f6b1faa5af9a

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read trailofbits--interpreting-culture-index at revision 2be6433b39e009cd2b6fd3be1895f6b1faa5af9a, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 43

references/anti-patterns.md

Common mistakes when interpreting Culture Index profiles. Avoiding these errors is as important as understanding the methodology itself.

<interpretation_mistakes>

Comparing absolute trait values between people

Wrong: "Dan has higher autonomy than Jim because his A is 8 vs 5"

Right: "Dan is +3 centiles from his arrow; Jim is +1 from his arrow"

Why it's wrong: The arrow position varies between surveys based on EU. An 8 with an arrow at 6 is only +2 from norm. A 5 with an arrow at 2 is +3 from norm.

Only exception: L and I use absolute values and CAN be compared directly.

Ignoring the red arrow position

The arrow is the population mean (50th percentile). All interpretation must be relative to it.

Wrong: "This person has low B (score of 3)"

Right: "This person's B is 2 centiles left of their arrow, indicating pronounced introversion"

Treating traits as "good" or "bad"

Culture Index measures behavioral traits, not value or capability.

Wrong:

  • "High D is good because they're detail-oriented"
  • "Low B is bad because they're not social"

Right:

  • "High D indicates strong attention to detail - fits roles requiring precision"
  • "Low B indicates preference for focused work - fits analytical roles"

There is no universally "good" profile. Fit depends on role, team, and context.

Using outdated data (18+ months old)

Job behaviors update as environment, leadership, or projects change. Stale data leads to wrong conclusions.

Best practice: Resurvey job behaviors every 6 months, especially after major role or leadership changes.

Over-indexing on a single trait

The pattern (relationship between traits) matters more than any individual dot.

Wrong: "They have High A, so they're a leader"

Right: "High A combined with High D suggests they can build new systems. High A with Low D suggests they'll drive fast but may miss details."

Not comparing Survey vs Job graphs

Missing burnout signals by only looking at one graph.

Always check:

  • Did the arrow move? (Stress/Frustration signal)
  • Did any dots flip sides? (Polarizing shift - flight risk)
  • What's the EU utilization? (>130% stress, <70% frustration)

Confusing traits with behaviors

Survey (top) = hardwired traits (who you ARE) Job (bottom) = adaptive behaviors (who you're TRYING TO BE)

Traits don't change. Behaviors can be modified temporarily - but at an energy cost.

"You can't send a duck to Eagle school." - You can train behaviors, not traits.

</interpretation_mistakes>

<application_mistakes>

Hiring for homogeneous teams

Hiring people who match the manager's profile creates blind spots.

Better approach: Build diverse teams with complementary traits. Every team needs Gas (High A), Brake (High D), and Glue (High B) in appropriate proportions.

Overloading on High A's

High A's are "the single hardest trait to employ." They work for "me, Inc." first.

Too many High A's = power struggles, lack of follow-through, no one to execute.

You are only RENTING High A's - they need a mutually beneficial partnership.

Neglecting the Brake (High D)

Every team needs quality control, risk management, and follow-through.

Without Brake: Erosion, mistakes, lawsuits, quality issues, things start but never finish.

Assuming job fit is permanent

People's roles evolve. Business needs change. What fit yesterday may not fit tomorrow.

Regular check-ins (resurvey every 6 months) catch misalignment before it becomes burnout or turnover.

Using Culture Index as the only data point

Culture Index is ONE tool among many. It measures behavioral traits, not:

  • Skills or competencies
  • Intelligence
  • Experience
  • Values
  • Motivation

Best practice: Use CI alongside interviews, references, skills assessments, and performance data.

</application_mistakes>

<communication_mistakes>

Sharing raw scores without context

Telling someone "Your A is 8" without explaining relative position is meaningless and potentially harmful.

Always communicate:

  • Position relative to arrow
  • What that means behaviorally
  • Why it's neither good nor bad
  • How it fits (or doesn't) with their role

Using CI to label or box people

CI shows tendencies, not destiny.

Wrong: "You're a Persuader, so you should only do sales"

Right: "Your pattern shows strengths in influence and relationship building. How do you use those in your current role?"

Comparing profiles publicly

Never compare individuals' profiles in group settings without their consent.

Discussing "Person A is more detail-oriented than Person B" creates hierarchy and judgment.

</communication_mistakes>

<red_flags>

Signals that suggest deeper investigation:

| Signal | What to Check | |--------|---------------| | EU 0-10 (avoidant response) | Was survey completed properly? Trust issues? | | All dots on or near arrow | Chameleon pattern - less than 0.57% of population. Verify validity. | | Job behaviors completely opposite of Survey | Imminent flight risk. What's causing this extreme modification? | | EU utilization > 150% | Severe stress. Immediate conversation needed. | | EU utilization < 50% | Severe disengagement. May have already mentally quit. | | D raised significantly in Job behaviors | Most common unsustainable stress pattern. Why do they feel they need to be so much more perfectionist? |

</red_flags>

Before finalizing any interpretation:

  • [ ] Did I use relative positions (distance from arrow)?
  • [ ] Did I avoid calling any trait "good" or "bad"?
  • [ ] Did I compare Survey vs Job if both available?
  • [ ] Did I calculate EU utilization?
  • [ ] Did I consider the full pattern, not just leading traits?
  • [ ] Is my data current (less than 18 months old)?
  • [ ] Did I note that CI is one data point among many?

<rationalization_table>

Common excuses that indicate you're about to make a CI interpretation mistake:

| Excuse | Reality | |--------|---------| | "Their A is higher so they're more autonomous" | Compare distance from arrow, not absolute values | | "This is a bad profile for leadership" | No bad profiles - fit depends on role and context | | "They need to change their C trait" | Survey traits are hardwired - change the environment instead | | "Low B means they're not a team player" | Low B means they prefer focused work - they can still collaborate | | "High D is always good for quality" | High D without other traits can mean paralysis and rigidity | | "They should be more like their manager" | Different profiles bring complementary strengths | | "This pattern can't do that job" | Patterns indicate tendencies, not hard limits | | "Their EU is fine, the job is the problem" | EU tells you about energy, not about job design | | "I remember their profile from last year" | Resurvey Job behaviors every 6 months - they change | | "The arrow doesn't matter, just look at the dots" | The arrow IS the reference point - dots mean nothing without it | | "L and I work the same as A, B, C, D" | L and I use absolute values, primary traits are relative | | "Survey and Job will be the same" | Survey = hardwired, Job = adaptive. They often differ. | | "High A means they're selfish" | High A means they're self-directed - not the same thing | | "We need all High A's on this high-growth team" | You're renting High A's, and they'll clash with each other | | "This hire looks good, skip the CI" | CI is one data point - use it WITH other assessments, not instead of |

</rationalization_table>