PR #15354: fix: add DeepSeek reasoning_content echo for tool-call messages (fixes #15353)

• Repository: NousResearch/hermes-agent
• Author: chen1749144759
• State: closed | merged: False
• Link: https://github.com/NousResearch/hermes-agent/pull/15354

Description (problem / solution / changelog)

Summary

DeepSeek V4 thinking mode requires reasoning_content on every assistant message that includes tool_calls. When missing, replay causes HTTP 400.

Closes #15353 Related: #14938 #14933 #15213

Changes

1. Merge DeepSeek into needs_tool_reasoning_echo check

In _copy_reasoning_content_for_api(), replaced the Kimi-only detection with a combined check covering:

• provider == "deepseek"
• "deepseek" in model (case-insensitive)
• api.deepseek.com base URL (custom provider)

This handles already-poisoned persisted sessions by injecting empty reasoning_content on replay.

2. Store reasoning_content on new tool-call messages

Added _needs_deepseek_tool_reasoning() helper method, wired into _build_assistant_message(). When a DeepSeek tool-call message is created without reasoning text (common for streaming tool-only turns), stores reasoning_content="" instead of omitting the field. Prevents future session poisoning at the source.

3. Fix _handle_max_iterations path

Added missing call to _copy_reasoning_content_for_api() in the max-iterations flush path. Previously only the main loop and flush_memories() had this call.

Test Plan

• Verified in state.db: new tool-call messages store reasoning_content
• Previously poisoned messages handled by replay fix
• Tested on Rocky Linux 9.7 with deepseek-v4-pro via custom provider

Diff

1 file changed: run_agent.py (+30, -3)

Changed files

• run_agent.py (modified, +30/-3)

---

PR #15407: fix: DeepSeek V4 thinking mode reasoning_content echo on tool-call messages

• Repository: NousResearch/hermes-agent
• Author: teknium1
• State: closed | merged: True
• Link: https://github.com/NousResearch/hermes-agent/pull/15407

Description (problem / solution / changelog)

Summary

DeepSeek V4 sessions with tool calls no longer die with 400 reasoning_content must be passed back. Fixes the creation path (new sessions are not poisoned) and the replay path (already-poisoned sessions recover).

Salvages @chen1749144759's #15354 with detection logic consolidated into helpers and a new regression test.

Root cause

DeepSeek V4 thinking mode requires reasoning_content on every assistant tool-call turn. Hermes had a Kimi-specific fallback in _copy_reasoning_content_for_api, but no DeepSeek coverage at creation time (_build_assistant_message) or replay time. Sessions got poisoned in state.db and every subsequent replay hit HTTP 400.

Changes

• run_agent.py:
  • _build_assistant_message pins reasoning_content="" on new tool-call turns when DeepSeek detected (prevents future poisoning).
  • _copy_reasoning_content_for_api padding now covers DeepSeek too (fixes poisoned history).
  • Extracted _needs_kimi_tool_reasoning() + _needs_deepseek_tool_reasoning() helpers — single source of truth, used by both the creation and replay paths.
  • Added missing _copy_reasoning_content_for_api() call in _handle_max_iterations() flush path (latent bug; was missing for Kimi too).
• tests/run_agent/test_deepseek_reasoning_content_echo.py: 21 tests covering all 3 DeepSeek signals (provider/model/host), poisoned replay, creation path, Kimi regression.
• scripts/release.py: AUTHOR_MAP entry for @chen1749144759.

Detection signals (DeepSeek)

• provider == "deepseek" (native)
• "deepseek" in model (custom-provider setups using deepseek model names)
• base_url host matches api.deepseek.com

Validation

Closes #15250, #15353. Supersedes #15228, #15354. Thanks @ruxme and @chen1749144759.

Changed files

• run_agent.py (modified, +44/-7)
• scripts/release.py (modified, +1/-0)
• tests/run_agent/test_deepseek_reasoning_content_echo.py (added, +213/-0)

---

PR #15446: fix(agent): comprehensive DeepSeek V4 support — context windows, thinking mode, reasoning replay

• Repository: NousResearch/hermes-agent
• Author: Tranquil-Flow
• State: open | merged: False
• Link: https://github.com/NousResearch/hermes-agent/pull/15446

Description (problem / solution / changelog)

Summary

Unifies 5 fragmented DeepSeek V4 PRs into a single cohesive implementation:

• Context windows: Add 1M entries for deepseek-v4-pro, deepseek-v4-flash, deepseek-chat, deepseek-reasoner (128K fallback preserved for older models)
• Thinking mode toggle: Plumb thinking.type and reasoning_effort for native DeepSeek API — maps effort values to DeepSeek's supported "high"/"max" pair, strips incompatible sampling params (temperature, top_p, etc.) when thinking is enabled
• reasoning_content replay: Inject reasoning_content="" on all assistant messages for DeepSeek replay, scoped to api.deepseek.com and OpenRouter deepseek/ prefix. Respects enabled: false to skip injection
• _extract_reasoning guards: Use isinstance(str) checks instead of truthy checks, preventing crashes on non-string reasoning values
• reasoning_content normalization: Preserve empty string "" in normalize_response (semantically valid for DeepSeek, was being dropped by truthy check)
• _handle_max_iterations: Add missing _copy_reasoning_content_for_api call so the max-iterations summary path doesn't produce 400s
• deepseek-chat preserved: deepseek-chat (the non-thinking alias) is NOT forced into thinking mode by default — only deepseek-v4-* and deepseek-reasoner models, or when the user explicitly opts in via reasoning_config

Test plan

• 34 tests across 5 test classes (context windows, thinking mode, replay, isinstance guards, normalization)
• Verified deepseek-chat does NOT force thinking mode by default
• Verified deepseek-chat CAN opt-in to thinking with explicit config
• Verified temperature stripping only happens when thinking is enabled
• Verified non-DeepSeek models are not affected by any of these changes
• Verified explicit reasoning_content is preserved (not overwritten with "")

Fixes #15353. Supersedes #14952, #14958, #15325, #15228, #15354.

Changed files

• agent/model_metadata.py (modified, +6/-2)
• agent/transports/chat_completions.py (modified, +33/-1)
• run_agent.py (modified, +38/-7)
• tests/agent/test_deepseek_v4.py (added, +322/-0)