Agent task: integrate agentplane with agent-registry, guardrails, model-router, and evidence ledger

[mdheller]

Agent-directed work package.

Target agent: Codex environment for SocioProphet/agentplane; Copilot coding agent if available.

Context:
agentplane is the execution control plane for governed bundles: validate, run, evidence, replay. It should consume agent-registry specs, guardrail policies, model-router policies, and model-governance ledger evidence without redefining those contracts.

Scope:

1. Inspect the live repository before editing.
2. Add docs/GOVERNED_AGENT_EXECUTION_INTEGRATION.md mapping AgentSpec, ToolGrant, GuardrailPolicy, ModelRoutingPolicy, PromotionRecord, and EvidenceRef into agentplane bundle execution.
3. Add an example governed agent bundle manifest that references agent-registry, guardrail-fabric, model-router, and model-governance-ledger artifacts.
4. Add repo.maturity.yaml using repo-maturity.v1 semantics.
5. Add or update validation targets for the example.

Acceptance criteria:

• Validation passes using the repository's Makefile or existing validation command.
• Docs clearly separate execution/replay from registry, routing, guardrails, and governance ledger ownership.
• The example shows evidence emission and replay references.
• No secrets, live tokens, credentials, model weights, or private datasets are committed.

Boundary:
One PR only. Do not touch unrelated repositories.

[mdheller]

@codex Please take this agentplane issue. Implement the scoped governed execution integration in the issue body and open one PR with validation evidence.

[mdheller]

@copilot Please take this agentplane issue if available. Implement the scoped governed execution integration and open one PR with validation evidence.

[chatgpt-codex-connector]

Summary

• Added a new integration design doc that scopes governed execution boundaries and maps external contracts (AgentSpec, ToolGrant, GuardrailPolicy, ModelRoutingPolicy, PromotionRecord, EvidenceRef) into agentplane consumption points, including replay/evidence boundaries. docs/GOVERNED_AGENT_EXECUTION_INTEGRATION.mdL1-L58
• Added a governed execution example manifest that references upstream agent-registry, guardrail-fabric, model-router, and model-governance-ledger artifacts, and explicitly includes both evidence emission and replay input references. examples/governed-execution/governed-agent-bundle.example.jsonL1-L102
• Added repo.maturity.yaml using repo-maturity.v1 semantics with maturity dimensions and explicit validation/governance evidence notes. repo.maturity.yamlL1-L20
• Added a new validator script for governed manifests to assert required integration references plus tool grants, evidence emission entries, and replay inputs. tools/validate_governed_manifest.pyL1-L173
• Updated the repository validation target so make validate now runs both existing execution timing validation and the new governed manifest validation. MakefileL1-L8

Testing

• ✅ make validate
• ✅ git status --short
• ✅ git commit -m "Add governed execution integration doc, manifest, and validation"
• ✅ nl -ba Makefile docs/GOVERNED_AGENT_EXECUTION_INTEGRATION.md examples/governed-execution/governed-agent-bundle.example.json repo.maturity.yaml tools/validate_governed_manifest.py | sed -n '1,260p'

Also completed:

• Committed changes on the current branch (c704027).
• Opened a PR via make_pr with title: “Integrate scoped governed execution references and validation into agentplane”.

View task →

[mdheller]

Triage — Integrate agentplane with agent-registry, guardrails, model-router, and evidence ledger: Architecture. Core integration hub. Very high scope — recommend breaking into four separate issues, one per dependency.