Issue Station
Sign In
Tools hereToolbox

openclaw - ✅(Solved) Fix CLI default --to routes to main while explicit --agent routes correctly (routing mismatch) [5 pull requests, 3 comments, 1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
openclaw/openclaw#56267•Fetched 2026-04-08 01:42:58
View on GitHub
Comments
3
Participants
1
Timeline
12
Reactions
0
Author
OscarlingOscarling
Participants
OscarlingOscarling
Timeline (top)
cross-referenced ×7commented ×3closed ×1locked ×1

Fix Action

Fixed

PR fix notes

PR #42: RH-T5-B01: escalate upstream issue and sync ledgers

Description (problem / solution / changelog)

Summary:

  • create upstream escalation record for RH-T5-B01
  • sync BACKLOG / DECISIONS / 验收清单 with A9 state
  • link local tracker issue #37 to upstream issue openclaw/openclaw#56267

Result:

  • blocker remains open
  • project enters wait-for-upstream-feedback mode with local guardrails still active

Links:

Changed files

  • BACKLOG.md (modified, +3/-3)
  • DECISIONS.md (modified, +18/-0)
  • design/validation/2026-03-28-role-hardening-rh-t5-b01-upstream-issue-escalation.md (added, +35/-0)
  • 验收清单.md (modified, +2/-2)

PR #43: RH-T5-B01: add upstream-feedback recheck trigger workflow

Description (problem / solution / changelog)

Summary:

  • add scripts/rh_t5_b01_upstream_recheck_bundle.sh to run A6+A7+A8 and output blocker_close_ready
  • add upstream feedback trigger card (event-driven)
  • add A10 baseline validation record
  • sync BACKLOG / DECISIONS / 验收清单

Baseline result:

  • blocker_close_ready=no
  • final_result=blocker_still_open

Links:

Changed files

  • BACKLOG.md (modified, +3/-2)
  • DECISIONS.md (modified, +25/-0)
  • design/validation/2026-03-28-rh-t5-b01-upstream-feedback-trigger-card-v1.md (added, +49/-0)
  • design/validation/2026-03-28-rh-t5-b01-upstream-recheck-bundle-baseline.md (added, +35/-0)
  • scripts/rh_t5_b01_upstream_recheck_bundle.sh (added, +112/-0)
  • 验收清单.md (modified, +6/-2)

PR #44: RH-T5-B01: add upstream feedback probe and A11 baseline

Description (problem / solution / changelog)

Summary:

  • add scripts/rh_t5_b01_upstream_feedback_probe.sh to classify upstream feedback trigger events
  • update trigger card to include probe-first workflow
  • add A11 baseline validation doc
  • harden premerge_check.sh to syntax-check all rh_t5_b01 scripts
  • sync BACKLOG / DECISIONS / 验收清单

A11 baseline result:

  • upstream_feedback_detected=no
  • local_tracking_issue_open=yes
  • next_event=waiting_upstream_feedback

Links:

Changed files

  • BACKLOG.md (modified, +2/-2)
  • DECISIONS.md (modified, +24/-0)
  • design/validation/2026-03-28-rh-t5-b01-upstream-feedback-probe-baseline.md (added, +35/-0)
  • design/validation/2026-03-28-rh-t5-b01-upstream-feedback-trigger-card-v1.md (modified, +11/-3)
  • scripts/premerge_check.sh (modified, +7/-0)
  • scripts/rh_t5_b01_upstream_feedback_probe.sh (added, +99/-0)
  • 验收清单.md (modified, +3/-3)

PR #45: RH-T5-B01: add event runner and A12 baseline

Description (problem / solution / changelog)

Summary:

  • add scripts/rh_t5_b01_event_runner.sh as single-entry event executor
  • update upstream feedback trigger card to prefer one-command flow
  • add A12 baseline validation doc
  • sync BACKLOG / DECISIONS / 验收清单

A12 baseline result:

  • next_event=waiting_upstream_feedback
  • action_taken=wait
  • action_result=waiting_upstream_feedback

Links:

Changed files

  • BACKLOG.md (modified, +2/-2)
  • DECISIONS.md (modified, +28/-0)
  • design/validation/2026-03-28-rh-t5-b01-event-runner-baseline.md (added, +30/-0)
  • design/validation/2026-03-28-rh-t5-b01-upstream-feedback-trigger-card-v1.md (modified, +11/-4)
  • scripts/rh_t5_b01_event_runner.sh (added, +135/-0)
  • 验收清单.md (modified, +3/-3)

PR #47: docs: record RH-T5-B01 A14 upstream follow-up action

Description (problem / solution / changelog)

Summary

  • post A14 follow-up comment to upstream issue openclaw/openclaw#56267
  • add validation record for upstream follow-up action
  • sync BACKLOG/DECISIONS/验收清单 with A14 state

Validation

  • python3 scripts/backlog_lint.py
  • python3 scripts/backlog_sync.py
  • bash scripts/premerge_check.sh

Notes

  • blocker RH-T5-B01 remains open
  • current event state stays waiting_upstream_feedback

Changed files

  • BACKLOG.md (modified, +2/-2)
  • DECISIONS.md (modified, +22/-0)
  • design/validation/2026-03-28-rh-t5-b01-upstream-followup-comment-validation.md (added, +27/-0)
  • 验收清单.md (modified, +2/-2)
RAW_BUFFERClick to expand / collapse

Summary: In OpenClaw CLI, default calls with --to route to main, while explicit --agent calls route correctly to the target agent. This creates a routing mismatch and can silently bypass configured default entry expectations.

Environment:

  • OpenClaw versions observed: 2026.3.12 and 2026.3.24 (same behavior)
  • Config default agent: steward (agents.list has steward default=true)
  • Route binding: steward bound to telegram account default

Reproduction:

  1. Set steward as default agent in config and verify via openclaw agents list --bindings --json.
  2. Run default path: openclaw agent --to +15550001111 --session-id probe-default --message test --json.
  3. Run explicit path: openclaw agent --agent steward --to +15550001111 --session-id probe-explicit --message test --json.
  4. Compare sessionKey in result.meta.systemPromptReport.

Observed:

  • default path sessionKey: agent:main:main
  • explicit path sessionKey: agent:steward:main
  • mismatch persists across repeated probes and controlled comparisons

Additional controlled evidence:

  • provider/model/profile controlled comparison: all calls used openai + gpt-5.4 + same profile, mismatch still persists
  • session-id controlled comparison: even with same session-id, default path routes to main while explicit path routes to steward
  • to-path probe: default no-to path stays on steward workspace; default with --to path resolves to workspace-main

Expected: When steward is configured as default entry and no explicit --agent is provided, CLI default --to behavior should be consistent with configured default routing expectations (or fail loudly if ambiguity is intentional).

Why this matters: This can silently misroute production tasks unless every caller enforces explicit --agent at wrapper level.

Request: Please confirm intended behavior for default --to routing and, if unintended, provide a fix or a documented config or flag to force default --to to honor configured default agent.

extent analysis

Fix Plan

To resolve the routing mismatch issue, we need to update the OpenClaw CLI to honor the configured default agent when no explicit --agent is provided.

Here are the concrete steps:

  • Update the openclaw agent command to check for the default agent in the config when no --agent flag is provided.
  • Use the default agent's routing binding if available.

Example code snippet:

def get_agent(config, args):
    if args.agent:
        return args.agent
    default_agent = config.get('default_agent')
    if default_agent:
        return default_agent
    return 'main'  # fallback to main agent

# Usage
config = {'default_agent': 'steward'}
args = argparse.Namespace(agent=None, to='+15550001111')
agent = get_agent(config, args)
print(agent)  # Output: steward

Verification

To verify the fix, run the reproduction steps again and compare the sessionKey in the result. The sessionKey should now be consistent with the configured default routing expectations.

Extra Tips

  • Make sure to update the documentation to reflect the new behavior.
  • Consider adding a flag to force the default --to behavior to honor the configured default agent, e.g., --use-default-agent.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#ssr#installation#tensor shape#request timeout

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

openclaw - ✅(Solved) Fix CLI default --to routes to main while explicit --agent routes correctly (routing mismatch) [5 pull requests, 3 comments, 1 participants]