TRUST - Trust Analysis

Input: $ARGUMENTS

Interpretations

Before executing, identify which interpretation matches the user’s input:

Interpretation 1 — Trust assessment: The user wants to understand the current state of trust in a relationship or organization — where trust is strong, where it’s broken, and what’s driving the gaps. Interpretation 2 — Trust repair: Trust has been damaged (by a specific event or pattern) and the user wants to understand what it would take to rebuild it — not platitudes about “rebuilding trust” but the specific mechanics. Interpretation 3 — Trust design: The user is building a team, partnership, or system and wants to understand how to establish trust from the start — what structures, behaviors, and norms create trust vs. erode it.

If ambiguous, ask: “I can help with assessing where trust stands now, repairing damaged trust, or designing trust into a new relationship or structure — which fits?” If clear from context, proceed with the matching interpretation.

Corruption Pre-Inoculation

Trust analysis easily becomes blame assignment. If your analysis clearly identifies one party as the trust violator and the other as the victim, check whether you’re seeing the full picture. Trust erosion is usually bidirectional, even when one event is more visible. Also: the user will present themselves as the trustworthy party. Test that.

Full protocol: _shared/corruption-pre-inoculation.md

Core Principles

1. Trust is not one thing. Trust has at least three independent dimensions: COMPETENCE (can they do it?), INTEGRITY (will they do what they said?), and BENEVOLENCE (do they care about my interests?). Someone can be trusted on one dimension and distrusted on another. Never say “I don’t trust them” — say WHICH dimension.

2. Trust is asymmetric. A trusts B differently than B trusts A. Each direction may differ on each dimension. A may trust B’s competence but not integrity, while B trusts A’s integrity but not competence. Map both directions.

3. Trust is built slowly and broken fast. One act of trustworthiness adds a small amount. One violation can erase years of trust. This asymmetry means trust maintenance is more important than trust building — the cost of violation is always higher than the benefit of one trustworthy act.

4. Vulnerability is the mechanism. Trust only exists where vulnerability exists. If there is no risk, there is no trust — just confidence. The question is always: what am I risking by trusting this person, and what evidence do I have that the risk is acceptable?

5. Systems create trust or destroy it. Trust is not purely personal. Organizational structures, incentive systems, information flows, and accountability mechanisms either make trust rational or make distrust rational. If the system rewards defection, individual trustworthiness is swimming upstream.

6. Stated trust is unreliable. People say “I trust you” for social reasons. Revealed trust — what they actually delegate, share, or risk — is the real signal. Watch behavior, not declarations.

Phase 1: TRUST MAPPING

Step 1: Define the Trust Landscape

CONTEXT: [relationship, team, organization, partnership]
PARTIES: [list all relevant parties — label as T1, T2, T3...]
TRIGGERING CONCERN: [what prompted this analysis — a specific event, pattern, or design question]

Step 2: Map Trust by Dimension and Direction

For EACH directional pair (T1→T2, T2→T1, etc.):

TRUST ASSESSMENT: [T1] → [T2]
================================

COMPETENCE TRUST: [HIGH / MEDIUM / LOW / UNKNOWN]
  Evidence: [what has T1 observed about T2's ability?]
  Key question: "Can T2 actually do what they're responsible for?"

INTEGRITY TRUST: [HIGH / MEDIUM / LOW / UNKNOWN]
  Evidence: [has T2 kept commitments? Been honest? Followed through?]
  Key question: "Will T2 do what they said they would do?"

BENEVOLENCE TRUST: [HIGH / MEDIUM / LOW / UNKNOWN]
  Evidence: [does T2 consider T1's interests? Show care beyond obligation?]
  Key question: "Does T2 actually care about my well-being?"

OVERALL TRUST POSTURE: [T1 toward T2]
  What T1 is willing to risk with T2: [delegations, disclosures, dependencies]
  What T1 withholds from T2: [things they won't delegate, share, or depend on]
  Gap between stated and revealed trust: [do they SAY they trust but ACT otherwise?]

Step 3: Identify Trust Asymmetries

ASYMMETRY MAP
=============

[A1] [dimension] trust: [T1]→[T2] = [level], [T2]→[T1] = [level]
  Impact: [what dysfunction does this asymmetry create?]
  Root cause: [why the asymmetry exists — different information, different experience, different standards]

[A2] ...

Phase 2: TRUST DYNAMICS

Step 4: Trace Trust History

TRUST TRAJECTORY
================

BUILDING EVENTS (actions that increased trust):
- [E1] [event] — increased [dimension] trust in [direction] because [reason]
- [E2] [event] — ...

EROSION EVENTS (actions that decreased trust):
- [E3] [event] — decreased [dimension] trust in [direction] because [reason]
- [E4] [event] — ...

VIOLATION EVENTS (actions that broke trust):
- [E5] [event] — broke [dimension] trust in [direction]
  Was it: [COMPETENCE failure / INTEGRITY breach / BENEVOLENCE betrayal]?
  Was it acknowledged? [yes / no / partially]
  Was it repaired? [yes / no / attempted but insufficient]

PATTERN: [is trust trending up, down, or oscillating? Why?]

Step 5: Analyze Structural Trust Factors

SYSTEM TRUST ANALYSIS
=====================

INCENTIVE ALIGNMENT:
- Do the parties' incentives align? [where yes, where no]
- Does the system reward trustworthy behavior? [or does defection pay?]

ACCOUNTABILITY STRUCTURES:
- Are commitments trackable? [or do they disappear into ambiguity?]
- Are violations visible? [or can they be hidden?]
- Are there consequences for trust violations? [or are they cost-free?]

INFORMATION FLOW:
- Is information shared symmetrically? [or does one party hoard?]
- Are decisions transparent? [or opaque?]
- Can parties verify claims? [or must they take on faith?]

POWER DYNAMICS:
- Who has more power? [and how does this affect trust?]
- Can the less powerful party hold the more powerful one accountable?
- Does power asymmetry make trust rational or irrational for the weaker party?

STRUCTURAL VERDICT: Does the SYSTEM make trust rational? [YES / PARTIALLY / NO]
If NO: [Individual trust-building efforts will fail without structural change. Identify what must change.]

Phase 3: TRUST ACTION PLAN

Step 6: Generate Trust Interventions

Based on the mapping and dynamics, identify specific interventions:

TRUST INTERVENTIONS
===================

FOR [dimension] TRUST, [direction]:

[I1] [specific action]
  Mechanism: [why this builds trust on this dimension]
  Cost: [what it requires — vulnerability, time, resources]
  Signal strength: [HIGH / MEDIUM / LOW — how much trust does this build?]
  Prerequisite: [what must be true for this to work]
  Timeline: [immediate / weeks / months]

[I2] [specific action]
  ...

Rules for interventions:

• Each must be SPECIFIC (not “be more transparent” but “share weekly project status including problems, not just progress”)
• Each must target a specific dimension and direction
• At least one must involve structural change, not just behavioral change
• At least one must involve vulnerability from the party seeking more trust
• Trust repair interventions must address the specific violation type (competence → demonstrate capability, integrity → over-deliver on commitments, benevolence → sacrifice self-interest visibly)

Step 7: Compile the Trust Analysis

TRUST ANALYSIS: [context]
==========================

TRUST STATE:
  Strongest trust: [T?]→[T?] on [dimension] — because [reason]
  Weakest trust: [T?]→[T?] on [dimension] — because [reason]
  Most damaging asymmetry: [A-number] — creating [specific dysfunction]

ROOT ISSUE: [Is this primarily a competence problem, integrity problem, benevolence problem, or structural problem?]

WHAT WOULD CHANGE THE MOST:
  Behavioral: [single highest-impact behavior change, with I-number]
  Structural: [single highest-impact system change]

WHAT WILL MAKE IT WORSE:
- [specific action to avoid — and why it erodes trust further]
- [specific action to avoid]

TRUST REPAIR SEQUENCE (if trust is damaged):
  1. [First — acknowledge the specific violation without deflection]
  2. [Second — demonstrate changed behavior on the violated dimension]
  3. [Third — sustain over time — trust rebuilds at the speed of consistent evidence]
  Expected timeline: [realistic estimate — not "it takes time" but how long and what milestones]

UNCOMFORTABLE TRUTH: [something the user probably doesn't want to hear about their own role in the trust dynamic]

CONFIDENCE: [HIGH / MEDIUM / LOW]
WEAKEST ASSUMPTION: [what part of this analysis is most likely wrong]

Failure Modes

Depth Scaling

Default: 2x. These are floors.

Pre-Completion Checklist

• Trust mapped by dimension (competence, integrity, benevolence) — not treated as monolithic
• Trust mapped bidirectionally — both directions analyzed
• Asymmetries identified with their dysfunction impact
• Structural factors analyzed — not just interpersonal
• Interventions are specific and actionable — not platitudes
• At least one intervention requires vulnerability from the party seeking trust
• Uncomfortable truth about the user’s own role included
• Trust repair timeline is realistic (months, not conversations)
• Depth floors met

Integration

• Use from: /conflict (trust damage often underlies conflict), /empth (understand why someone distrusts)
• Routes to: /conflict (if trust erosion has created active conflict), /persua (if rebuilding trust requires a persuasive case), /empth (if one party’s trust posture is poorly understood)
• Differs from: /conflict (conflict maps disagreement structure; trust maps reliability expectations), /empth (empathy models someone’s full perspective; trust focuses specifically on reliability dimensions)
• Complementary: /empth (understand WHY someone distrusts), /conflict (resolve conflicts that trust erosion created), /aw (stress-test your trust assessment)