Skip to content
Michael Rishi Forrester

Agentic Covenants

Agentic Charter Matrix

Who is allowed to put this agent into the world, under what authority, accountable to what policy, with what retirement criteria?

Filter

All 15Agent charterDomain charterOrganizational charterIdentityAuthorizationBlast radiusApproval gatingSupply chain

Layers, by what they can actually enforce

  • Agent charter — advisory

    What the agent's own charter declares. Scope it claims for itself.

  • Domain charter — deterministic-when-invoked

    The domain or team that operates the agent, and the policy they set.

  • Organizational charter — deterministic-and-external

    Authority above the operator. Organizational policy the operator cannot waive.

Showing 15 of 15.

  • Identity at Organizational charter

    AI Acceptable Use Policy names categories of agents allowed to exist. AI Governance Council holds authority to create new agent classes. Director of AI Workforce Transformation is the named org-wide owner.

  • Identity at Domain charter

    Domain leadership signs charter authorizing a class of agents. Names roles permitted to create agents and the escalation path.

  • Identity at Agent charter

    Named human owner accountable for the agent's actions. Backup owner identified. Identity claims tied to charter via registered agent identifier.

  • Authorization at Organizational charter

    AI Risk Appetite Statement defines hard prohibitions and change-control process for evolving scope policy.

  • Authorization at Domain charter

    Per-class scope, inherits org hard prohibitions, adds domain-specific restrictions.

  • Authorization at Agent charter

    Specific authorized scope. Named tools, MCP servers, environments, max-blast-radius. Scope expansion requires re-signature.

  • Blast radius at Organizational charter

    Org-wide risk tier taxonomy (1 read-only, 2 scoped writes, 3 destructive, 4 production-critical) with damage caps and matching control requirements.

  • Blast radius at Domain charter

    Inherits tier taxonomy, defines which tiers the domain is authorized to operate, defines failure-mode reviews per tier.

  • Blast radius at Agent charter

    Specific risk tier, specific damage cap (records-per-session, USD-per-day, forbidden ops), conditions for tier downgrade or retirement.

  • Approval gating at Organizational charter

    AI Governance Council approves new agent classes, ratifies risk-tier policy. Member roles named (CISO, Chief AI Officer, GC, Privacy, Domain Leads). Quorum and voting rules defined.

  • Approval gating at Domain charter

    Domain authority approves charters within its domain. Multi-party signature for Tier 3+. Charter amendments require the same process as originals.

  • Approval gating at Agent charter

    Charter signed by owner, domain authority, and (Tier 3+) security review. Charter identifies approver of every subsequent scope change. Annual review cadence. Emergency revocation conditions specified.

  • Supply chain at Organizational charter

    Org-wide allowlist of approved foundation models. Org-wide policy on MCP server approval, third-party dependency approval, vendor risk assessment integrated with procurement.

  • Supply chain at Domain charter

    Inherits org-wide approved-model list, adds domain-specific restrictions (e.g., "no models that train on user data", "SOC 2 Type II vendors only"). Approves or denies MCP servers for the domain.

  • Supply chain at Agent charter

    Specific dependencies declared. Named foundation model and version, named MCP servers (with hashes from Covenants L2-C5), named base images, named tool versions. Dependency changes require charter amendment.