Plan: Explicit ACP delegation override (2.6.3)

[lodyai]

Summary

• Drafts the ExecPlan for roadmap entry 2.6.3, docs/execplans/2-6-3-explicit-acp-delegation-override.md, covering the operator-facing acp.delegation configuration knob, the wiring that promotes CapabilityPolicy::MaskAndDeny to production for agent.mode = "acp", the per-family parametrization of the initialize masker and runtime denylist, configuration-validation diagnostics, the per-session trust-boundary audit event, and the reservation of JSON-RPC application error code -32002.
• Signposts ADR 006 (capability dispositions), ADR 008 (operator-override trust boundary precedent), docs/podbot-design.md §Execution flow, and the Corbusier conformance §ACP delegation section.
• Enumerates nine staged milestones (A through I), each with its own validation gate and natural commit boundary, plus a Decision log recording the choices made during planning (per-family override granularity, four-array MethodDenylist parametrization, dual audit-event emission, host-only CLI flag surface, reserved-but-unused -32002 code, proptest over the narrowing invariant).

The plan itself is draft-status; implementation begins only after this plan is approved. make markdownlint and make nixie both pass on the new document. The plan does not modify any code; it adds a single new document.

Test plan

• Approve the plan (or request revisions) before any code lands.
• Implementation Stages A–I land on the same branch, each with its own gate-passing commit.
• make check-fmt, make lint, make test succeed after each milestone.
• coderabbit review --agent reports zero unresolved concerns at the final milestone.
• docs/podbot-roadmap.md Step 2.6.3 checkbox is marked done at the end.

References

• Roadmap entry: docs/podbot-roadmap.md §2.6.3 (Add the explicit ACP delegation override).
• ExecPlan: docs/execplans/2-6-3-explicit-acp-delegation-override.md.
• Upstream Steps: docs/execplans/2-6-1-intercept-acp-initialization.md, docs/execplans/2-6-2-runtime-denylist.md.
• Architectural Decision Records: ADR 006 (validate surface and capability dispositions), ADR 008 (operator-override trust-boundary precedent).
• Lody session: https://lody.ai/leynos/sessions/6dd80e6b-1c56-4d39-b6c0-c89aefb7bc1e

🤖 Generated with Claude Code

Summary by Sourcery

Documentation:

• Introduce a detailed ExecPlan for roadmap step 2.6.3 describing the explicit ACP delegation override configuration, its trust-boundary implications, staged implementation milestones, and validation/acceptance criteria.

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: c53f9854-63c1-4f5a-b35c-cbced3b2c501

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch 2-6-3-explicit-acp-delegation-override

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[sourcery-ai]

Reviewer's Guide

Adds a new, detailed ExecPlan markdown document for roadmap step 2.6.3 that specifies how to implement an explicit ACP delegation override, including configuration, engine wiring, validation, auditing, testing, and documentation work, without changing any production code yet.

Sequence diagram for ACP delegation override validation and session audit

sequenceDiagram
    actor Operator
    participant CLI as CLI_HostArgs
    participant ConfigLoader
    participant AppConfig
    participant Validator as validate_agent_config
    participant Tracing as tracing_warn
    participant Engine as ProtocolSession

    Operator->>CLI: pass --acp-delegation allow_fs
    CLI->>ConfigLoader: build ConfigLoadOptions
    ConfigLoader->>AppConfig: deserialize to AppConfig(acp.delegation)

    AppConfig->>Validator: validate_agent_config()
    alt delegation != Disabled and agent.mode == Acp
        Validator->>Tracing: warn target=podbot::acp::policy
        Tracing-->>Validator: recorded
        Validator-->>AppConfig: Ok
    else delegation != Disabled and agent.mode != Acp
        Validator-->>AppConfig: ConfigError::InvalidValue
    end

    Opt session_start
        AppConfig->>Engine: build ProtocolSessionOptions
        Engine->>Engine: emit_trust_boundary_audit_event()
        Engine->>Tracing: warn target=podbot::acp::policy
        Tracing-->>Engine: recorded
    end

Loading

File-Level Changes

Change	Details	Files
Introduce a comprehensive execution plan for implementing the explicit ACP delegation override for ACP-hosted agents.	Add a new ExecPlan document describing the purpose and scope of step 2.6.3, including constraints, tolerances, risks, and success criteria for the ACP delegation override feature. Define the planned acp.delegation configuration knob, its allowed values, validation rules, and how it will interact with agent.mode and AppConfig. Specify the design and wiring for a pure delegation selector (AcpDelegationOverride, AcpCapabilityMask, AcpSessionPolicy) feeding CapabilityPolicy, MethodDenylist, and the initialise-frame masker. Plan semantic config validation and two audit events (at config load and per session) to surface trust-boundary changes via tracing::warn! on the podbot::acp::policy target. Describe how MethodDenylist and the initialise masker will be parametrised to stay consistent under overrides and how JSON-RPC error code -32002 will be reserved for future override-required diagnostics. Lay out staged implementation work (Stages A–I) including unit tests, rstest-bdd behavioural tests, proptest invariants, CLI plumbing, and documentation updates, along with acceptance and quality criteria.	docs/execplans/2-6-3-explicit-acp-delegation-override.md

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.