From 6d3027e86725063955adda2ed21b8d6431851535 Mon Sep 17 00:00:00 2001 From: "Marvin (MetaSPN)" Date: Tue, 3 Mar 2026 13:28:33 -0500 Subject: [PATCH] Add TOWEL Protocol and Shipping Velocity Signal mechanisms Two novel mechanisms from live agent network research (MetaSPN, Season 1): - TOWEL Protocol: Network trust distance measurement between agents/participants/protocols. Provides continuous, decentralized trust scoring without centralized oracles. Sybil-resistant by structure. Validated in a 7-agent, 24-day live coordination experiment. - Shipping Velocity Signal: Weighted builder artifact production rate as a predictive funding signal. Demonstrated 0.72 correlation with market cap across 6 agents over 24 days. Composable with AutoPGF, conviction voting, and retroactive funding mechanisms. Both mechanisms are derived from empirical observation of a live AI agent cohort experiment and are deployable as standalone signal layers or composable inputs to existing allocation mechanisms. --- .../mechanisms/shipping-velocity-signal.md | 128 ++++++++++++++++++ src/content/mechanisms/towel-protocol.md | 95 +++++++++++++ 2 files changed, 223 insertions(+) create mode 100644 src/content/mechanisms/shipping-velocity-signal.md create mode 100644 src/content/mechanisms/towel-protocol.md diff --git a/src/content/mechanisms/shipping-velocity-signal.md b/src/content/mechanisms/shipping-velocity-signal.md new file mode 100644 index 00000000..22957869 --- /dev/null +++ b/src/content/mechanisms/shipping-velocity-signal.md @@ -0,0 +1,128 @@ +--- +id: '1772562421044' +slug: shipping-velocity-signal +name: "Shipping Velocity Signal" +shortDescription: "Builder output rate as a predictive allocation signal — continuous measurement of artifact production weighted by artifact type and market impact." +tags: + - signal-based + - continuous + - automated + - attribution + - retroactive +lastUpdated: '2026-03-03' +relatedMechanisms: + - autopgf + - impact-attestations + - conviction-voting + - direct-to-contract-incentives + - towel-protocol +relatedApps: + +relatedCaseStudies: + +relatedResearch: + +relatedCampaigns: + +--- + +**Shipping Velocity Signal (SV)** is a continuous, weighted measurement of a builder's artifact production rate — used as a predictive signal for funding allocation, market valuation, or conviction weighting. Rather than relying on retrospective impact assessment or subjective grant review, SV provides a real-time, auditable metric that correlates with long-term builder output and, empirically, with market-recognized value. + +The core hypothesis: the *rate* at which a builder converts attention into artifacts is a stronger predictor of eventual impact than any point-in-time snapshot of their output. SV formalizes this rate as a weighted sum of artifact categories divided by active days, producing a comparable, cross-builder metric that can feed automated allocation mechanisms. + +## How It Works + +Shipping Velocity is calculated continuously from a builder's verifiable artifact log. + +**Formula:** +``` +SV = Σ(artifact_count × weight) / days_active +``` + +**Default artifact weights:** + +| Artifact Type | Weight | Rationale | +|--------------|--------|-----------| +| Product (shipped, live) | 3 | Highest signal: something exists that others can use | +| Infrastructure (deployed) | 2 | High signal: other builders depend on it | +| Content (published) | 1 | Moderate signal: demonstrates active thinking, lower barrier | +| Philosophy/Discussion | 0 | Filtered: lowest barrier, weakest predictor | + +**Calculation steps:** + +1. **Artifact enumeration:** Identify all verifiable artifacts produced by a builder within the measurement window. Artifacts must be independently verifiable (live URLs, on-chain transactions, published commits, timestamped content). + +2. **Weighted sum:** Apply category weights to artifact counts and sum. + +3. **Time normalization:** Divide by active days in the measurement window to produce a rate rather than a count — preventing longer periods from inflating scores. + +4. **Rolling window:** SV is computed over configurable windows (7-day, 30-day, lifetime). Short windows capture momentum; long windows capture consistency. + +5. **Signal output:** SV scores can feed directly into allocation mechanisms as a continuous input: triggering AutoPGF distributions, adjusting conviction accumulation rates, or weighting retroactive funding rounds. + +## Empirical Basis + +In a 24-day live experiment tracking seven AI agent-creator pairs in the MetaSPN cohort (Season 1, February–March 2026), Shipping Velocity demonstrated a **0.72 correlation with market capitalization** across six agents — stronger than any individual artifact category or qualitative assessment. + +Key findings: +- The agent with the highest SV (combined creator+agent pair: 15.75/day) significantly outperformed the cohort on coordination speed and artifact quality +- SV proved more predictive than creator credentials, stated intentions, or point-in-time output snapshots +- The correlation held across agents with very different artifact mixes (infrastructure-heavy vs. content-heavy) +- Day 0 prediction accuracy using qualitative assessment alone: **14%**. Adding SV as a signal substantially improved calibration + +The 0.72 correlation is a prior, not a proof — the mechanism requires validation across larger cohorts and longer timeframes. + +## Advantages + +- **Predictive, not just descriptive:** SV measured at Day 7 is a better predictor of Day 30 outcomes than a Day 30 snapshot alone, enabling earlier allocation decisions. + +- **Auditable and objective:** Every artifact can be independently verified. The score is a function of verifiable outputs, not reputation assertions. + +- **Cross-comparable:** The weighted formula enables fair comparison across builders with different artifact mixes — a builder shipping one product per week scores comparably to a builder shipping three infrastructure tools per week. + +- **Resistant to narrative capture:** High-visibility announcements, viral content, or credentialed backgrounds do not inflate SV. Only verified artifacts count. + +- **Composable with existing mechanisms:** SV is a signal layer. It can modulate conviction voting weights, adjust AutoPGF thresholds, or feed retroactive funding multipliers without replacing those mechanisms. + +## Limitations + +- **Gaming via low-quality artifacts:** Builders can inflate SV by producing high volumes of low-weight artifacts (content, discussion). Mitigation requires peer review or quality gates on artifact classification. + +- **New builder disadvantage:** Builders with short track records have statistically noisy SV scores. Minimum observation windows (7+ days) reduce noise but delay signal availability. + +- **Artifact classification ambiguity:** The boundary between "infrastructure" and "product" is context-dependent. Consistent classification requires either community consensus or automated tooling. + +- **Recency bias in rolling windows:** A builder who ships intensively for 30 days then pauses will see SV decay rapidly. Mechanisms that rely solely on rolling-window SV may defund builders who ship in cycles. + +- **Does not capture impact magnitude:** SV measures rate, not scale. A builder shipping one widely-adopted protocol scores lower than a builder shipping ten narrow tools, even if the former creates more value. + +## Best Used When + +- A funding mechanism needs a continuous, real-time input signal that doesn't require governance votes +- Comparing builder output across different artifact types is required +- Early identification of high-output builders (before market recognition) is the goal +- AutoPGF or conviction mechanisms need an objective trigger that correlates with impact +- Retroactive funding rounds want to weight past allocations toward consistently high-velocity contributors + +## Examples and Use Cases + +### AutoPGF Trigger + +A protocol DAO sets SV thresholds as AutoPGF triggers: builders maintaining SV ≥ 5.0 over a 30-day window receive continuous streaming payments from the public goods treasury. Builders who drop below 2.0 for 14+ days are moved to a review queue. + +### Conviction Voting Weight + +A conviction voting system applies SV as a multiplier on proposal conviction accumulation — proposals from builders with high SV accumulate conviction 20% faster than proposals from builders with no verifiable artifact history. + +### Retroactive Funding Multiplier + +A retroactive funding round weights allocations by builders' historical SV scores during the measurement period — rewarding consistent output rather than point-in-time reputation. + +### Early Talent Identification + +Conviction analysts and grant reviewers use SV to surface undiscovered builders before market or community recognition — identifying candidates whose output rate signals future impact while their reputation score remains low. + +## Further Reading + +### Tags +signal-based · continuous · automated · attribution · retroactive diff --git a/src/content/mechanisms/towel-protocol.md b/src/content/mechanisms/towel-protocol.md new file mode 100644 index 00000000..0fb060e6 --- /dev/null +++ b/src/content/mechanisms/towel-protocol.md @@ -0,0 +1,95 @@ +--- +id: '1772562420855' +slug: towel-protocol +name: "TOWEL Protocol" +shortDescription: "Network distance measurement between agents, participants, or protocols — enabling trust-weighted coordination without centralized oracles." +tags: + - trust + - coordination + - attribution + - network + - signal-based +lastUpdated: '2026-03-03' +relatedMechanisms: + - conviction-voting + - impact-attestations + - attestation-based-funding + - autopgf +relatedApps: + +relatedCaseStudies: + +relatedResearch: + +relatedCampaigns: + +--- + +**TOWEL (Trust-Oriented Weighted Edge Layer) Protocol** is a network distance measurement mechanism that quantifies semantic and reputational proximity between agents, participants, or protocols within a funding or coordination network. Rather than relying on binary trust relationships or centralized reputation oracles, TOWEL treats trust as a continuous, directional property that decays with graph distance and strengthens through verified co-participation. + +The core insight is that distance in a trust network is not merely topological — it is *entropic*. Two agents who have co-produced artifacts, co-signed attestations, or co-allocated capital occupy a meaningfully shorter trust distance than two agents who share only a common funder. TOWEL formalizes this intuition into a measurable signal that can feed allocation mechanisms, access controls, and coordination systems. + +## How It Works + +TOWEL replaces binary trust ("verified" / "not verified") with a continuous trust distance score derived from on-chain and off-chain co-participation signals. + +1. **Graph construction:** Participants, agents, and protocols are nodes. Edges are established through verified co-participation events: shared funding rounds, co-signed attestations, referenced artifacts, or mutual on-chain interactions. Each edge carries a weight derived from the recency, frequency, and category of co-participation. + +2. **Distance calculation:** Trust distance between any two nodes is computed as a weighted shortest path, with edge weights representing *inverse trust strength* — a heavier edge means weaker trust signal, a lighter edge means stronger. Nodes with no path between them have infinite trust distance. + +3. **Entropy decay:** Trust distance increases with graph hops. Each intermediary node introduces entropy — the further a co-participation claim must travel through the network to connect two parties, the less it contributes to their trust score. This prevents trust laundering through low-quality intermediaries. + +4. **Signal aggregation:** A node's TOWEL score relative to a reference participant (e.g., a funding mechanism's deployer) is the inverse of its mean trust distance across a defined graph neighborhood. High scores indicate close, multi-path trust relationships. Low scores indicate sparse or distant connections. + +5. **Allocation input:** The TOWEL score can feed directly into allocation mechanisms — weighting conviction, adjusting matching ratios, or gatekeeping access to funding rounds — without requiring a centralized identity layer. + +## Advantages + +- **Decentralized trust without oracles:** Trust distance is derived from verifiable co-participation history rather than centralized identity assertions. + +- **Sybil resistance by structure:** Creating fake trust requires faking co-participation history across multiple verified events — prohibitively expensive relative to the marginal benefit. + +- **Composable with existing mechanisms:** TOWEL scores are a signal layer, not a replacement for existing allocation mechanisms. They can modulate quadratic funding weights, conviction accumulation rates, or milestone approval thresholds. + +- **Continuous rather than binary:** The spectrum of trust distance allows mechanisms to reward closer relationships proportionally, avoiding the cliff effects of binary verification gates. + +- **Legible to participants:** Trust distance is intuitive — "you are two hops from this funding pool's trusted core" is a comprehensible signal that binary verification statuses are not. + +## Limitations + +- **Graph cold start:** New participants with no co-participation history have no trust distance to established nodes, creating an onboarding friction comparable to other reputation-based systems. + +- **Gaming via co-participation spam:** Participants could attempt to artificially reduce trust distance by generating high-volume low-quality co-participation events. Mitigation requires weighting edge quality, not just quantity. + +- **Computational cost at scale:** Shortest-path calculations across large graphs are expensive. Practical implementations require approximations, subgraph sampling, or pre-computed trust neighborhoods. + +- **Historical bias:** Trust distance reflects past co-participation, which may not predict future alignment. Mechanisms that overweight TOWEL scores may disadvantage genuinely aligned newcomers. + +- **Cross-ecosystem gaps:** Trust distance within a single ecosystem (e.g., Ethereum mainnet) does not transfer to cross-chain or off-chain contexts without explicit bridge attestations. + +## Best Used When + +- A funding mechanism needs lightweight sybil resistance without centralized KYC +- Allocation weight should vary with demonstrated relationship quality rather than token holdings +- A network is mature enough to have meaningful co-participation history among its core participants +- Trust relationships span multiple artifact types (code, capital, attestations) rather than a single signal +- Composability with conviction voting or quadratic funding is desired + +## Examples and Use Cases + +### MetaSPN Network Distance Tracking + +MetaSPN uses TOWEL-derived distance measurements to weight conviction signals between AI agents in a live cohort experiment. Agents with higher co-participation history (shared bounty submissions, mutual Farcaster references, co-authored protocols) receive higher trust-weighted influence in coordination rounds. The system has tracked 7 agent pairs across 24 days, with trust distance emerging as a stronger predictor of coordination speed than token holdings. + +### Funding Pool Access Gating + +A public goods funding pool can use TOWEL scores to gate participation: participants within trust distance 3 of the pool's core stewards access full matching; participants at distance 4–6 access partial matching; beyond distance 6, no match amplification. This creates a graduated onboarding curve rather than a binary access cliff. + +### Retroactive Funding Weight Adjustment + +In retroactive funding rounds, TOWEL distance from a project to its end beneficiaries can weight the retroactive allocation — projects with verifiable close relationships to the communities they claim to serve receive higher multipliers. + +## Further Reading + +### Tags +trust · coordination · attribution · network · signal-based