Your Trust Score Is Real. The Signals Behind It Are Manufactured.

Helps identify when a skill's trust reputation is built on coordinated
social manipulation rather than genuine community validation.

Problem

Trust in agent marketplaces flows through social signals: upvotes, downloads,
comments, and follow counts. These signals are valuable precisely because they
aggregate distributed judgment — when thousands of independent users find a
skill useful and safe, their collective assessment carries real information.

The assumption of independence is the attack surface. A coordinated network
of accounts can manufacture the appearance of distributed consensus. A skill
with 500 upvotes from a bot network looks identical to a skill with 500
upvotes from 500 independent developers. The marketplace's reputation system
cannot distinguish manufactured trust from earned trust — and neither can
most agents that rely on reputation as a trust signal.

Social trust manipulation is the third pillar of the trust attack surface,
alongside technical attacks (code injection) and structural attacks (supply
chain compromise). It is the most scalable: a well-constructed sockpuppet
network can manufacture trust faster than any code-level auditing can catch
it, and the manufactured trust persists long after the network is dismantled.

Legitimate skills earn trust gradually, from a diverse user base, with
engagement patterns that correlate with actual skill utility. Manipulated
skills earn trust in coordinated bursts, from accounts with suspicious
creation patterns, with engagement that does not correlate with usage or
outcomes.

What This Detects

This detector examines social trust integrity across five dimensions:

1. 1. Engagement velocity anomalies — Does the skill's vote/download

trajectory show natural growth curves, or coordinated burst patterns? Organic trust accumulates gradually; manufactured trust arrives in synchronized bursts that are statistically distinguishable from random arrival processes

1. 2. Account cohort analysis — Do the skill's early upvoters share

creation dates, activity patterns, or cross-voting behavior that suggests coordinated rather than independent operation? Sockpuppet networks leave structural fingerprints in how accounts relate to each other

1. 3. Engagement-to-utility correlation — Do social signals correlate with

actual skill usage metrics? High upvotes on skills with low actual install rates, or high engagement from users who only interact with one publisher's skills, are signals of manufactured rather than genuine trust

1. 4. Cross-publisher coordination — Do multiple publishers in a marketplace

show correlated voting patterns, where their respective supporter networks upvote each other's skills at rates that exceed random baseline? Coordinated mutual-support networks amplify manufactured trust across multiple accounts simultaneously

1. 5. Review authenticity signals — Do comments and reviews on the skill

show the linguistic diversity and specificity expected from independent users, or do they share vocabulary, complaint patterns, or phrasing that suggests template-generated or coordinated content?

How to Use

Input: Provide one of:

• - A skill identifier to assess the authenticity of its trust signals
• A publisher account to analyze for coordinated network membership
• A set of skills to assess for cross-publisher coordination patterns

Output: A manipulation detection report containing:

• - Engagement velocity analysis (organic vs. burst pattern)
• Account cohort fingerprint assessment
• Engagement-to-utility correlation score
• Cross-publisher coordination indicators
• Review authenticity assessment
• Manipulation verdict: AUTHENTIC / SUSPICIOUS / COORDINATED / MANUFACTURED

Example

Input: Assess social trust integrity for ai-assistant-toolkit publisher

CODEBLOCK0

Related Tools

• - clone-farm-detector — Detects content-level cloning for reputation gaming;

social-trust-manipulation-detector catches social-level gaming that can occur even with original, non-cloned content

• - publisher-identity-verifier — Verifies publisher identity integrity;

sockpuppet networks may impersonate multiple independent publishers when they are controlled by a single actor

• - trust-velocity-calculator — Quantifies trust decay from update velocity;

manufactured trust does not decay the same way as earned trust and creates distorted velocity measurements

• - blast-radius-estimator — Estimates propagation impact if a skill is

compromised; skills with manufactured trust may have artificially high install counts that misrepresent actual blast radius

Limitations

Social trust manipulation detection depends on access to engagement metadata
(account creation dates, cross-voting patterns, install counts) that many
marketplaces do not expose through public APIs. Where metadata is limited,
only velocity analysis and review text assessment are available, which reduces
detection confidence. Burst voting patterns can result from legitimate causes:
coordinated community launches, press coverage, or featured placement can all
produce rapid engagement that resembles manufactured trust. The account cohort
analysis relies on observable fingerprints and will miss well-resourced
adversaries who age accounts and vary patterns. This tool identifies social
trust signals that warrant investigation — it does not confirm manipulation,
which requires access to platform-level data that only marketplace operators
can verify.