Skip to content

docs(ollama): add prominent warnings about /v1 breaking tool calling#21437

Closed
lanelupo wants to merge 1 commit intoopenclaw:mainfrom
lanelupo:docs/ollama-tool-calling-warning
Closed

docs(ollama): add prominent warnings about /v1 breaking tool calling#21437
lanelupo wants to merge 1 commit intoopenclaw:mainfrom
lanelupo:docs/ollama-tool-calling-warning

Conversation

@lanelupo
Copy link
Copy Markdown

@lanelupo lanelupo commented Feb 19, 2026

Summary

  • Problem: Users running Ollama on a remote host naturally configure baseUrl: "http://host:11434/v1" with api: "openai-responses", which silently breaks tool calling
  • Why it matters: Model outputs raw JSON as plain text instead of executing tools—no error surfaced, confusing UX
  • What changed: Added prominent <Warning> callouts at top of doc, in custom URL section, and in legacy mode section
  • What did NOT change: No code changes, docs-only

Change Type (select all)

  • Bug fix
  • Feature
  • Refactor
  • Docs
  • Security hardening
  • Chore/infra

Scope (select all touched areas)

  • Gateway / orchestration
  • Skills / tool execution
  • Auth / tokens
  • Memory / storage
  • Integrations
  • API / contracts
  • UI / DX
  • CI/CD / infra

Linked Issue/PR

User-visible / Behavior Changes

None (docs-only). Users will now see clear warnings before misconfiguring Ollama.

Security Impact (required)

  • New permissions/capabilities? No
  • Secrets/tokens handling changed? No
  • New/changed network calls? No
  • Command/tool execution surface changed? No
  • Data access scope changed? No

Repro + Verification

Environment

  • OS: Linux (Ubuntu)
  • Runtime/container: N/A (docs change)
  • Model/provider: Ollama
  • Integration/channel: N/A
  • Relevant config: N/A

Steps

  1. View docs/providers/ollama.md before change
  2. Note warning about /v1 is buried in Advanced section
  3. View docs/providers/ollama.md after change
  4. Note prominent warning at top and in custom URL section

Expected

  • Clear warning visible before user configures remote Ollama incorrectly

Actual

  • Warning now appears at top of doc and in relevant config sections

Evidence

  • Diff shows <Warning> callouts added in three locations
  • Renders correctly in Mintlify (standard component)

Human Verification (required)

  • Verified scenarios: Read through doc flow as a new user setting up remote Ollama
  • Edge cases checked: Ensured warning appears before any config example that could mislead
  • What I did NOT verify: Live Mintlify rendering (no local Mintlify instance)

Compatibility / Migration

  • Backward compatible? Yes
  • Config/env changes? No
  • Migration needed? No

Failure Recovery (if this breaks)

  • How to disable/revert: Revert commit
  • Files/config to restore: docs/providers/ollama.md
  • Known bad symptoms: N/A (docs-only)

Risks and Mitigations

None. Docs-only change adding warnings.

- Add top-level warning for remote Ollama users about /v1 trap
- Clarify custom base URL section: show api: 'ollama' explicitly
- Add inline warning about not adding /v1 to URL
- Make Legacy OpenAI-Compatible Mode warning more prominent
- Clarify that tool calling does not work with openai-completions

Fixes openclaw#21243
@openclaw-barnacle openclaw-barnacle bot added docs Improvements or additions to documentation size: XS labels Feb 19, 2026
@vincentkoc
Copy link
Copy Markdown
Member

A fix was pushed in #29204.

@vincentkoc vincentkoc closed this Feb 27, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

docs Improvements or additions to documentation size: XS

Projects

None yet

Development

Successfully merging this pull request may close these issues.

docs: warn more clearly that openai-responses/openai-completions breaks tool calling for remote Ollama

2 participants