chore(adr): make agent execution reproducible

[yordis]

• Agent behavior needs stable, reviewable semantics across revisions and Sessions.
• Model access and credential custody need explicit boundaries so reproducibility and security claims remain meaningful.

[cursor]

PR Summary

Medium Risk
Documentation-only, but it sets normative boundaries for revisions, immutable sessions, and credential custody that later auth and execution code must implement without silent fallback.

Overview
Defines reproducible agent execution and hosted model access in architecture docs only—no runtime code.

ADR#0025 drops the stable runtime field and model_default in favor of AgentConfiguration pins for agent_implementation (exact version, typed config, digests) and primary/auxiliary ModelSelection. Sessions are described as copying those pins into an immutable SessionExecutionPlan and resolving only external facts (routes, bindings, hosting) per ADR#0031, with revised revision-boundary tables, charter/learned classes, and examples.

ADR#0031 (new) normatively requires exact implementation and model pins at admission, a canonical immutable plan at SessionStarted, AgentImplementationAdapter + Ready attestation, ExecutionAttempt evidence, fail-closed admission (no Session-time upgrades), and typed protobuf oneof modeling (including OpenClaw/Codex/Claude classifications).

ADR#0032 (new) separates ResolvedModelRoute, ModelProviderConnection, CredentialBinding, and attempt-scoped **ModelAccessGrant**s; routes resolve before run; credentials stay out of agent/implementation state; a single model access service brokers upstream calls via attested supervisors with no native fallback on auth/model/provider errors.

docs/adr/index.md indexes ADR#0031 and #32.

^{Reviewed by Cursor Bugbot for commit 396bbea. Bugbot is set up for automated code reviews on this repo. Configure here.}

[coderabbitai]

Warning

Review limit reached

@yordis, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 26 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 49866a95-e5d5-4b27-92d1-b4fdaae7e8a1

📥 Commits

Reviewing files that changed from the base of the PR and between c870d13 and 396bbea.

📒 Files selected for processing (4)

• docs/adr/0025-agent-definition-data-ownership.md
• docs/adr/0031-agent-implementation-and-session-plan.md
• docs/adr/0032-model-route-and-credential-binding.md
• docs/adr/index.md

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch yordis/adr-harness-credentials

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 0d792a2. Configure here.}