fix: suppress tool-call planning content

[0xCheetah1]

Summary

• suppress assistant content when the same response includes tool_calls, preventing provider planning text from being forwarded to chat surfaces
• preserve tool_calls so tool execution still works normally
• add a regression test for Kimi-style planning content alongside tool calls
• type the plugin reload config used by the current fork baseline

Testing

• Verified on VPS with moonshot/kimi-k2.6 through OpenClaw Telegram: old installer leaked planning text; patched branch returned clean final answers
• npm run build on patched branch

Summary by CodeRabbit

• Improvements

  • Assistant responses now suppress planning/thinking prose whenever tool calls are present, while preserving tool-call data.
  • Added support for an optional reload configuration (noopPrefixes) in plugin definitions.

• Tests

  • Updated tests to cover tool-call response behavior and upstream server responses.

• Documentation

  • Fixed markdown table formatting in several skill docs for consistent rendering.

• Style

  • Minor code formatting tweaks with no functional changes.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: b2db58b1-0306-4dfb-afcf-3a38eae5d7c5

📥 Commits

Reviewing files that changed from the base of the PR and between 4a4c9ee and 6488348.

📒 Files selected for processing (6)

• skills/clawrouter/SKILL.md
• skills/imagegen/SKILL.md
• src/index.ts
• src/proxy.tool-forwarding.test.ts
• src/proxy.ts
• src/types.ts

✅ Files skipped from review due to trivial changes (4)

• skills/clawrouter/SKILL.md
• src/index.ts
• src/types.ts
• skills/imagegen/SKILL.md

🚧 Files skipped from review as they are similar to previous changes (2)

• src/proxy.tool-forwarding.test.ts
• src/proxy.ts

---

📝 Walkthrough

Walkthrough

Adds an optional reload.noopPrefixes to plugin types, reformats a small expression in src/index.ts, changes proxy streaming and non-streaming response handling to suppress assistant prose when tool_calls are present, refactors proxy tests to use a configurable upstream response, and updates two skill README tables' formatting.

Changes

Cohort / File(s)	Summary
Plugin Types & small index tweak src/types.ts, src/index.ts	Added optional reload?: { noopPrefixes?: string[] } to OpenClawPluginDefinition. Minor single-line reformat of toolWidth calculation in src/index.ts (no behavior change).
Proxy Response Handling src/proxy.ts	Streaming and non-streaming response post-processing now detect per-choice tool_calls and suppress assistant prose by setting emitted delta.content / message.content to "" when non-empty tool calls exist; preserves structured tool_calls in output; conditional re-serialization applied.
Tool-forwarding Tests src/proxy.tool-forwarding.test.ts	Refactored tests to use a mutable upstreamResponse reset in beforeEach; added test asserting proxy returns HTTP 200, preserves tool_calls, and suppresses assistant message.content when upstream signals finish_reason: "tool_calls".
Documentation formatting skills/clawrouter/SKILL.md, skills/imagegen/SKILL.md	Adjusted Markdown table layouts and minor spacing in two SKILL.md files; only formatting changes, no content or behavioral changes.

Sequence Diagram(s)

sequenceDiagram
  participant Client
  participant Proxy
  participant Upstream
  Client->>Proxy: request
  Proxy->>Upstream: forward request
  Upstream-->>Proxy: response (choices with/without tool_calls)
  alt tool_calls present
    Proxy->>Proxy: for each choice, detect tool_calls → set content = ""
    Proxy-->>Client: stream tool_calls chunks + empty assistant content
  else no tool_calls
    Proxy->>Proxy: strip "thinking" tokens from content
    Proxy-->>Client: stream assistant content
  end

Loading

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title directly summarizes the main change: suppressing planning/thinking content in tool-call responses. This aligns with the core objective stated in PR objectives.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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 and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

src/proxy.tool-forwarding.test.ts (1)

118-179: Add a streaming-path companion regression test.

Line [118]-[179] correctly covers non-streaming behavior. Add a stream: true case that asserts empty delta.content with preserved tool_calls so SSE behavior is also locked in.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/proxy.tool-forwarding.test.ts` around lines 118 - 179, Add a companion
streaming test that mirrors the non-streaming case: create a new it(...) e.g.
"suppresses assistant content when upstream returns tool_calls (streaming)", set
upstreamResponse the same way, POST to `${proxy.baseUrl}/v1/chat/completions`
with stream: true and same messages/tools, consume the SSE response and assert
that the streamed assistant delta events contain an empty delta.content and
include the tool_calls payload (preserved) — i.e. verify each relevant SSE
chunk's parsed delta has content === "" and tool_calls length === 1 (and that
the final finish event is handled); reuse the same identifiers
(upstreamResponse, proxy.baseUrl, endpoint /v1/chat/completions) so the test
structure matches the existing non-streaming test.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@src/proxy.tool-forwarding.test.ts`:
- Around line 118-179: Add a companion streaming test that mirrors the
non-streaming case: create a new it(...) e.g. "suppresses assistant content when
upstream returns tool_calls (streaming)", set upstreamResponse the same way,
POST to `${proxy.baseUrl}/v1/chat/completions` with stream: true and same
messages/tools, consume the SSE response and assert that the streamed assistant
delta events contain an empty delta.content and include the tool_calls payload
(preserved) — i.e. verify each relevant SSE chunk's parsed delta has content ===
"" and tool_calls length === 1 (and that the final finish event is handled);
reuse the same identifiers (upstreamResponse, proxy.baseUrl, endpoint
/v1/chat/completions) so the test structure matches the existing non-streaming
test.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 43046a9b-b7f2-4132-b6f7-779a7cebc116

📥 Commits

Reviewing files that changed from the base of the PR and between 3c00b45 and 926a610.

📒 Files selected for processing (4)

• src/index.ts
• src/proxy.tool-forwarding.test.ts
• src/proxy.ts
• src/types.ts