Skip to content

fix(agents): preserve compatible CLI session runtime pins#92476

Merged
vincentkoc merged 2 commits into
openclaw:mainfrom
yu-xin-c:codex/session-runtime-compat-provider
Jun 13, 2026
Merged

fix(agents): preserve compatible CLI session runtime pins#92476
vincentkoc merged 2 commits into
openclaw:mainfrom
yu-xin-c:codex/session-runtime-compat-provider

Conversation

@yu-xin-c

@yu-xin-c yu-xin-c commented Jun 12, 2026

Copy link
Copy Markdown
Contributor

Summary

  • Preserve persisted agentRuntimeOverride values for CLI runtimes when the runtime is registered to serve the selected provider.
  • Keep stale CLI runtime pins isolated: a claude-cli session pin still does not leak into unrelated providers such as openai.
  • Pass runtime config through the reply/follow-up harness override resolver so setup-registered CLI runtime aliases can be validated.

Real behavior proof

Behavior or issue addressed:
Persisted session runtime pins were only restored for the hard-coded openai + codex pair. A session pinned to a setup-registered CLI runtime such as claude-cli would be dropped even when the selected provider was the matching model provider (anthropic). This could force the fallback/restore path away from the CLI runtime that originally owned the session. The fix keeps the existing stale-pin guard by restoring a CLI runtime only when isCliRuntimeAliasForProvider confirms that the runtime serves the selected provider.

Real environment tested:
Local OpenClaw checkout on macOS, branch codex/session-runtime-compat-provider, Node/Vitest through repository scripts.

Exact steps or command run after this patch:

node scripts/run-vitest.mjs run src/auto-reply/reply/agent-runner-execution.test.ts -t "resolveSessionRuntimeOverrideForProvider" --reporter verbose
node scripts/run-vitest.mjs run src/auto-reply/reply/agent-runner-execution.test.ts --reporter default
./node_modules/.bin/oxfmt --check --threads=1 src/auto-reply/reply/agent-runner-execution.ts src/auto-reply/reply/agent-runner-execution.test.ts src/auto-reply/reply/dispatch-from-config.ts src/auto-reply/reply/followup-runner.ts
node --import tsx --input-type=module --eval "<resolveSessionRuntimeOverrideForProvider live proof>"

Evidence after fix:

  • Focused Vitest: 2 passed | 188 skipped for resolveSessionRuntimeOverrideForProvider.
  • Full related file: agent-runner-execution.test.ts (190 tests) passed.
  • oxfmt: All matched files use the correct format.
  • Live proof output:
{
  "anthropic": "claude-cli",
  "openai": null
}

Observed result after fix:
With a setup-registered claude-cli runtime bound to anthropic, resolveSessionRuntimeOverrideForProvider({ provider: "anthropic", agentRuntimeOverride: "claude-cli" }) now returns "claude-cli"; the same persisted runtime pin still returns undefined for provider: "openai".

What was not tested:
End-to-end external CLI execution; this patch is limited to runtime override resolution and the reply/follow-up call sites that pass config into it.

@openclaw-barnacle openclaw-barnacle Bot added proof: supplied External PR includes structured after-fix real behavior proof. size: S labels Jun 12, 2026
@clawsweeper

clawsweeper Bot commented Jun 12, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Codex review: needs maintainer review before merge. Reviewed June 12, 2026, 11:05 AM ET / 15:05 UTC.

Summary
The PR restores configured provider-compatible CLI runtime pins across reply, fallback, and follow-up resolution paths and adds matching-versus-stale regression coverage.

PR surface: Source +12, Tests +47. Total +59 across 4 files.

Reproducibility: yes. Current main has a high-confidence source reproduction: an anthropic session persisted with agentRuntimeOverride set to claude-cli reaches a resolver that accepts only openai/codex and therefore drops the compatible pin.

Review metrics: none identified.

Merge readiness
Overall: 🐚 platinum hermit
Proof: 🐚 platinum hermit
Patch quality: 🐚 platinum hermit
Result: ready for maintainer review.

Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch.

Rank-up moves:

  • none.

Next step before merge

  • The proposed patch is correct and needs only normal maintainer review and completion of the remaining required checks, not a separate repair PR.

Security
Cleared: The patch changes internal runtime selection and tests without altering dependencies, secrets, permissions, downloads, package resolution, or other supply-chain surfaces.

Review details

Best possible solution:

Land the registry-backed restoration rule so persisted CLI sessions retain their owning runtime only when the configured backend explicitly serves the selected provider.

Do we have a high-confidence way to reproduce the issue?

Yes. Current main has a high-confidence source reproduction: an anthropic session persisted with agentRuntimeOverride set to claude-cli reaches a resolver that accepts only openai/codex and therefore drops the compatible pin.

Is this the best way to solve the issue?

Yes. Reusing the canonical config-aware CLI backend binding is narrower and more maintainable than adding more hard-coded provider/runtime pairs, and the patch applies it consistently to the relevant reply and follow-up paths.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against e728957989f6.

Label changes

Label changes:

  • add P2: The regression can redirect resumed sessions away from their configured CLI runtime, but it affects a bounded set of CLI-backend users and has a focused fix.
  • add proof: sufficient: Contributor real behavior proof is sufficient. The PR provides after-fix live resolver output showing both compatible restoration and cross-provider rejection, supplemented by focused and full related-file tests.
  • add rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🐚 platinum hermit and patch quality is 🐚 platinum hermit.
  • add status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR provides after-fix live resolver output showing both compatible restoration and cross-provider rejection, supplemented by focused and full related-file tests.

Label justifications:

  • P2: The regression can redirect resumed sessions away from their configured CLI runtime, but it affects a bounded set of CLI-backend users and has a focused fix.
  • rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🐚 platinum hermit and patch quality is 🐚 platinum hermit.
  • status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (live_output): The PR provides after-fix live resolver output showing both compatible restoration and cross-provider rejection, supplemented by focused and full related-file tests.
  • proof: sufficient: Contributor real behavior proof is sufficient. The PR provides after-fix live resolver output showing both compatible restoration and cross-provider rejection, supplemented by focused and full related-file tests.
Evidence reviewed

PR surface:

Source +12, Tests +47. Total +59 across 4 files.

View PR surface stats
Area Files Added Removed Net
Source 3 13 1 +12
Tests 1 47 0 +47
Docs 0 0 0 0
Config 0 0 0 0
Generated 0 0 0 0
Other 0 0 0 0
Total 4 60 1 +59

What I checked:

Likely related people:

  • shakkernerd: Authored the recent commit that introduced the restrictive session runtime override resolver during agent-runtime internalization. (role: introduced current behavior; confidence: high; commits: f7ee25291a4b; files: src/auto-reply/reply/agent-runner-execution.ts)
  • steipete: Authored the merged agent-runtime internalization work and substantial surrounding CLI backend and runtime-routing history. (role: feature owner; confidence: high; commits: 995b055003c7, 859eb0666282; files: src/agents/cli-backends.ts, src/agents/model-runtime-aliases.ts, src/auto-reply/reply/agent-runner-execution.ts)
  • vincentkoc: Recently maintained the shared CLI backend and model-runtime alias surfaces, including canonical Codex runtime/auth routing. (role: recent adjacent contributor; confidence: medium; commits: 8c802aa68351, 859eb0666282; files: src/agents/cli-backends.ts, src/agents/model-runtime-aliases.ts)
What the crustacean ranks mean
  • 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
  • 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
  • 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
  • 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
  • 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
  • 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
  • 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

How this review workflow works
  • ClawSweeper keeps one durable marker-backed review comment per issue or PR.
  • Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
  • A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
  • PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
  • Maintainers can also comment @clawsweeper review to request a fresh review only.
  • Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
  • Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
  • Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

@clawsweeper clawsweeper Bot added proof: sufficient ClawSweeper judged the real behavior proof convincing. rating: 🐚 platinum hermit Good normal PR readiness with ordinary maintainer review expected. status: 👀 ready for maintainer look ClawSweeper has no concrete contributor-facing blocker left for this PR. P2 Normal backlog priority with limited blast radius. labels Jun 12, 2026
@vincentkoc
vincentkoc force-pushed the codex/session-runtime-compat-provider branch from 593de2a to 303b2f7 Compare June 13, 2026 22:50
@openclaw-barnacle openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label Jun 13, 2026
vincentkoc
vincentkoc approved these changes Jun 13, 2026
View reviewed changes

@vincentkoc vincentkoc left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Verified for maintainer merge.

What changed during maintainer review:

  • Kept the PR's provider-scoped CLI runtime pin fix.
  • Added the same compatible CLI-session guard to preflight compaction and memory flush, so anthropic sessions pinned to claude-cli do not enter embedded maintenance paths.

Proof:

  • node scripts/run-vitest.mjs src/auto-reply/reply/agent-runner-execution.test.ts src/auto-reply/reply/agent-runner-memory.test.ts src/agents/model-runtime-aliases.test.ts --maxWorkers=1 passed.
  • Autoreview rerun clean: no accepted/actionable findings.
  • Crabbox AWS cbx_44400b494e97 / coral-prawn, run run_69dd43475e39: OPENCLAW_CHECK_CHANGED_REMOTE_CHILD=1 OPENCLAW_CHANGED_LANES_RAW_SYNC=1 CI=1 corepack pnpm check:changed passed.
  • Exact PR head 303b2f794f6c01fcf21b62b27c536b5f6eceb421: 135 success / 24 skipped / 1 neutral / 0 failures / 0 pending.

@vincentkoc
vincentkoc merged commit 9974641 into openclaw:main Jun 13, 2026
164 of 165 checks passed
@clawsweeper clawsweeper Bot mentioned this pull request Jun 15, 2026
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@vincentkoc vincentkoc vincentkoc approved these changes

Assignees

No one assigned

Labels

P2 Normal backlog priority with limited blast radius. proof: supplied External PR includes structured after-fix real behavior proof. rating: 🐚 platinum hermit Good normal PR readiness with ordinary maintainer review expected. size: S status: 👀 ready for maintainer look ClawSweeper has no concrete contributor-facing blocker left for this PR.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@yu-xin-c @vincentkoc