fix(telegram): show full think levels for live-discovered Ollama models by xydttsw · Pull Request #93882 · openclaw/openclaw

xydttsw · 2026-06-17T03:00:26Z

AI-assisted PR. Implemented and verified with Claude (Claude Code).

Summary

What: Telegram /think inline menu was showing only default, off for live-discovered Ollama thinking models (e.g., ollama/glm-5.2:cloud), even though /think <level> commands like /think medium were accepted and functional.
Root cause: The menu resolution path used buildConfiguredModelCatalog() which reads only statically-configured models. Models discovered at runtime (via the {"ollama/*":{}} pattern or automatic Ollama endpoint discovery) were missing from this catalog, so their reasoning capability was unavailable to the thinking profile resolver.
Fix: Load the full runtime model catalog via loadModelCatalog() when building the command arg menu, ensuring all discovered models and their capabilities are available to the thinking level choice resolver.

Real behavior proof (required for external PRs)

Behavior addressed:
Telegram /think menu now shows the correct thinking levels for live-discovered Ollama models instead of only default, off.

Real environment tested:
OpenClaw dev build with Ollama extension on macOS 26.5.1 with ollama/glm-5.2:cloud model discovered via Ollama endpoint.

Exact steps or command run after this patch:

# 1. Configure Ollama provider with runtime discovery
openclaw config set providers '{"ollama":{"baseUrl":"http://localhost:11434"}}'
# 2. Switch to an Ollama model that supports thinking/openclaw reason
/model ollama/glm-5.2:cloud
# 3. Open the /think inline menu in Telegram
/think

Before-fix evidence:

/model ollama/glm-5.2:cloud
/think
> Current thinking level: high
> Options: default, off
/think medium
> Thinking level set to medium.

The /think menu shows only default, off even though /think medium is accepted — the menu doesn't reflect the actual available levels.

After-fix evidence:

/model ollama/glm-5.2:cloud
/think
> Current thinking level: high
> Options: default, off, low, medium, high, max

The /think menu now shows all thinking levels supported by the Ollama model, consistent with what /think commands actually accept.

Observed result after the fix:
After loading the runtime model catalog, resolveCommandArgMenu has access to the full model reasoning capability metadata, enabling it to present all supported thinking levels for live-discovered models.

What was not tested:

Non-Telegram channels (TUI, Discord) have separate menu resolution paths.
Edge cases with models that lack reasoning capability entirely (should still show default, off correctly).

Tests and validation

✓ extension build succeeds (TypeScript compilation)
✓ Existing bot-native-commands tests pass

Risk / scope

User visible behavior change: Yes (positive — menu now matches reality)
Configuration/environment changes: No
Security/credentials/network changes: No
Highest risk point: None — the change is passing a richer catalog to an existing function parameter
Mitigation: The loadModelCatalog() function is already used elsewhere in the same module for resolveTelegramDefaultThinkingLevel

clawsweeper · 2026-06-17T03:22:01Z

ClawSweeper review: did not complete due to Codex infrastructure failure. Reviewed June 16, 2026, 11:57 PM ET / 03:57 UTC.

Summary
Review failed before ClawSweeper could summarize the requested change.

PR surface: Source +5. Total +5 across 1 file.

Reproducibility: unclear. The review failed before ClawSweeper could establish a reproduction path.

Review metrics: none identified.

Merge readiness
Not assessed.
Failure reason: invalid structured output.

This is a ClawSweeper/Codex infrastructure failure, not a PR readiness or patch-quality verdict.
Keep any merge decision on the normal maintainer review path until ClawSweeper can complete a fresh review.

Risk before merge

[P1] No close action taken because the review did not complete.

Maintainer options:

Decide the mitigation before merge
Retry the Codex review after fixing the execution failure.
Pause or close
Do not merge this PR until maintainers decide whether the risk is worth taking.

Next step before merge

Review did not complete, so no work-lane recommendation was made.

Review details

Best possible solution:

Retry the Codex review after fixing the execution failure.

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

Unclear. The review failed before ClawSweeper could establish a reproduction path.

Is this the best way to solve the issue?

Unclear. Retry the review first so ClawSweeper can evaluate the actual issue and fix direction.

AGENTS.md: unclear because the file could not be read completely.

Codex review notes: model internal, reasoning high; reviewed against 0c909ea97f20.

Label changes

Label changes:

remove proof: sufficient: Current real behavior proof status is not_applicable, not sufficient.
remove mantis: telegram-visible-proof: Current Telegram visible-proof status is not_needed.
remove P2: Current review triage priority is none.
remove rating: 🐚 platinum hermit: Current review failed before PR readiness was assessed, so no rating label should remain.
remove status: 👀 ready for maintainer look: Current PR status no longer selects a status label.

Evidence reviewed

PR surface:

Source +5. Total +5 across 1 file.

View PR surface stats

Area	Files	Added	Net
Source	1	5	+5
Tests	0	0	0
Docs	0	0	0
Config	0	0	0
Generated	0	0	0
Other	0	0	0
Total	1	5	+5

What I checked:

failure reason: invalid structured output.
codex failure detail: Codex review failed for this PR with exit 0 and wrote invalid JSON or schema-invalid output to /home/runner/work/clawsweeper/clawsweeper/artifacts/event/codex/93882.json: decision.rootCauseCluster.currentItemRelationship fixed_by_candidate requires a canonical PR.
codex stderr: No stderr captured.
codex stdout: [],"mergeRiskLabels":[],"mergeRiskOptions":[],"reviewMetrics":[{"label":"Runtime catalog lookup","value":"1 conditional lookup added","reason":"The lookup is limited to native command menus that need unresolved model-aware choices, which is the right scope for the reported Telegram /think menu mismatch."}],"labelJustifications":[{"label":"P2","reason":"This is a normal-priority Telegram command-menu bugfix for live-discovered Ollama reasoning models with limited blast radius."}],"itemCategory":"bug","reproductionStatus":"source_reproducible","reproductionConfidence":"high","requiresNewFeature":false,"requiresNewConfigOption":false,"requiresProductDecision":false,"reproductionAssessment":"Yes. Source inspection reproduces the mismatch: current main passes only the configured catalog to the Telegram /think menu resolver, while the linked issue's live Ollama reasoning capabilit.

Likely related people:

unknown: Codex failed before it could trace repository history. (role: review did not complete; confidence: low)

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.

The /think menu in Telegram was only showing 'default, off' for live-discovered Ollama models because resolveCommandArgMenu used buildConfiguredModelCatalog (static config) instead of the full runtime model catalog. Runtime-discovered models (via '{"ollama/*":{}}' pattern) were missing from the catalog, so their 'reasoning' capability wasn't available to the thinking profile resolver. Fix: load the full runtime catalog via loadModelCatalog() when building the command arg menu, giving the thinking level resolver access to all discovered models and their capabilities. Fixes openclaw#93835

xydttsw · 2026-06-19T09:25:23Z

close all pr

openclaw-barnacle Bot added channel: telegram Channel integration: telegram size: XS triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. labels Jun 17, 2026

clawsweeper Bot had a problem deploying to qa-live-shared June 17, 2026 03:30 Failure

openclaw-barnacle Bot removed the triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. label Jun 17, 2026

openclaw-barnacle Bot added the triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. label Jun 17, 2026

Alix-007 mentioned this pull request Jun 17, 2026

fix(ollama): show full thinking levels for discovered reasoning models #93946

Closed

clawsweeper Bot mentioned this pull request Jun 17, 2026

fix(ollama): show full thinking levels for live-discovered models in /think menu #93872

Closed

openclaw-barnacle Bot added proof: supplied External PR includes structured after-fix real behavior proof. and removed triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. labels Jun 17, 2026

xydttsw force-pushed the fix/think-menu-ollama branch from d8d085b to fd6b90f Compare June 17, 2026 08:21

xydttsw closed this Jun 17, 2026

xydttsw reopened this Jun 17, 2026

clawsweeper Bot mentioned this pull request Jun 19, 2026

fix(ollama): treat undefined reasoning as potentially reasoning-capable #94882

Closed

7 tasks

xydttsw closed this Jun 19, 2026

clawsweeper Bot mentioned this pull request Jun 19, 2026

fix(ollama): show all thinking levels in /think menu for dynamically-discovered models #93864

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

fix(telegram): show full think levels for live-discovered Ollama models#93882

fix(telegram): show full think levels for live-discovered Ollama models#93882
xydttsw wants to merge 1 commit into
openclaw:mainfrom
xydttsw:fix/think-menu-ollama

xydttsw commented Jun 17, 2026 •

edited

Loading

Uh oh!

clawsweeper Bot commented Jun 17, 2026 •

edited

Loading

Uh oh!

xydttsw commented Jun 19, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Uh oh!

Conversation

xydttsw commented Jun 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Real behavior proof (required for external PRs)

Tests and validation

Risk / scope

Uh oh!

clawsweeper Bot commented Jun 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

xydttsw commented Jun 19, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

xydttsw commented Jun 17, 2026 •

edited

Loading

clawsweeper Bot commented Jun 17, 2026 •

edited

Loading