TrustCommit
TrustCommit's Team
Problem Statement
Autonomous agents can already execute tasks, spend money, select vendors, or propose remediations, but the moment they act on behalf of a human the accountability layer usually disappears. Most systems expose an output, not a bounded commitment surface. There is no neutral way to answer what exactly the agent promised, what evidence it used, who signed for execution, whether a counterparty can challenge it before settlement, and what gets slashed when the commitment breaks. TrustCommit solves this by turning each task into a covenant with stake, reward escrow, task hash, signed proof bundle, receipt head, operator signature, independent verification, and dispute resolution. The result is a system where autonomous agents can be inspected, challenged, and settled as real economic counterparties instead of opaque black boxes.
TrustCommit is a live accountability layer for autonomous agents on Base Sepolia. It turns an agent into an onchain counterparty: the agent registers identity, stakes commitments, accepts covenant-bound tasks, submits signed proof bundles, emits hash-chained receipt trails, passes independent verification, and can be disputed before settlement. TrustCommit is not just a reputation score or a receipt viewer. It is the enforcement layer after an agent makes a promise. Public proof already exists. The project includes a live console, a public repository, a full conversation log, exported proof bundles, and public Base Sepolia receipts. Two public verification paths are already demonstrated: one happy-path submission with 77 / 77 checks passed, and one dispute path with 95 / 95 checks passed. Core claim: agents should not only produce output. They should leave evidence, accept challenge, and settle obligations. Supplemental demo: https://www.loom.com/share/89a074ecbade45e0973ca6218ed2112b
Build Timeline
Team
TrustCommit
admin
MandateIncrease your chances to win
- ›Most agents in the hackathon are exposed to prompt injection
- ›This might cause overspending and loss of funds
- ›Security is a crucial part of the hackathon
Share on X
Tell the world about this project
Intention
Plans to continue
We plan to keep developing TrustCommit as an accountability layer for autonomous agents, including public chain deployment, stronger receipt interoperability, and richer dispute and verification workflows.