Your personal checks and balances.

Set the rules once. Change them whenever you want.

Your Self-Sovereign Civic Profile can hold the jurisdictions, neighbourhoods, issues, political interests, values, ethics, policy preferences, financial concerns and family priorities you choose to provide.

You decide the evidence requirements, alert thresholds and trade-offs. Those private inputs guide your agent. They are not a public political profile or a catalogue for sponsors to buy.

What the agent watches

Agendas, minutes, bills, bylaws, budgets, notices, recorded votes, public statements and official responses can reveal activity that matters to you. The Civic Graph connects each event to the institution, jurisdiction and authority involved.

Source availability and coverage remain visible. Silence in the record is not evidence that nothing happened.

The Civic Agent loop

Government acts → Civic Graph detects → Civic Agent compares → Relevance, alignment, conflict or uncertainty → Alert or authorized action → Participant remains informed.

A comparison can show potential alignment, potential conflict, a trade-off, insufficient evidence or a contradictory record. It should explain the source and the rule involved rather than silently turn a broad value into a political position.

Choose the level of authority

Watch

Follow the institutions and sources you select.

Filter

Surface material changes against your configured thresholds.

Explain

Summarize the evidence and show what remains uncertain.

Compare

Compare public action with your private rules and preferences.

Prepare

Draft a response for you to review, change or reject.

Execute standing instructions

Act only within a narrow, explicit, revocable instruction you previously created. Otherwise, return the issue to you for a decision.

A standing instruction is specific

For example: “When my municipality proposes an annual property-tax increase above X%, record concern and notify me.” You supply the threshold, scope, permitted action and duration. A configured value is not permission to invent a position on an unrelated issue.

Every delegated response must carry the label Executed under participant standing instruction and identify the applicable instruction. You remain informed and can change or end future authority.

Keep the two response states distinct

Human confirmed means the participant confirmed that particular response. Participant-authorized agent execution means the agent executed an existing standing instruction. It is authenticated participant authority, but it is not a fresh human confirmation.

Research, public interfaces and response records must preserve that distinction. A Mirror Response is non-binding; an official submission requires a separate authorized process and the relevant government's rules.

Begin with the founding network

The agent architecture is under development. The current registration form records your locality, interests and contact details; it does not connect your VerusID or activate monitoring, research access or delegated authority.

Build my Civic Agent