ROB-2197 Allow passing custom sys prompt param in Holmes chat endpoint

[mershal]

No description provided.

[coderabbitai]

Walkthrough

Threads an optional additional_system_prompt from the API model through server.py into build_chat_messages and add_or_update_system_prompt; when provided it is appended to the constructed system prompt for new and existing conversations and integrated into message history.

Changes

Cohort / File(s)	Summary
Core conversation prompt handling holmes/core/conversations.py	Added additional_system_prompt: Optional[str] = None parameter to add_or_update_system_prompt and build_chat_messages; forwards and appends the provided string into system prompt construction and when integrating system prompts into conversation history (affects context-window/tool-path logic).
API model update holmes/core/models.py	Added additional_system_prompt: Optional[str] = None field to ChatRequestBaseModel.
Server integration server.py	Forwards chat_request.additional_system_prompt into build_chat_messages in chat(); no other control-flow changes.

Sequence Diagram(s)

sequenceDiagram
  autonumber
  participant Client
  participant Server
  participant Builder as build_chat_messages
  participant SysUpd as add_or_update_system_prompt
  participant LLM

  Client->>Server: POST /chat (ChatRequest incl. additional_system_prompt)
  Server->>Builder: build_chat_messages(ask, history, ai, config, global_instructions, additional_system_prompt)
  Builder->>SysUpd: add_or_update_system_prompt(history, ai, config, additional_system_prompt)
  Note over SysUpd: If provided, append additional_system_prompt to the system prompt text
  SysUpd-->>Builder: updated conversation messages (with system prompt)
  Builder-->>Server: final messages
  Server->>LLM: invoke LLM with messages/tools
  LLM-->>Server: response
  Server-->>Client: chat reply

Loading

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• Share more logic for system prompt templating #727 — Implements passing extra system-prompt content through the request/flow (same feature of appending additional system prompt text).

Suggested reviewers

• mainred

Pre-merge checks and finishing touches

❌ Failed checks (2 warnings)

Check name	Status	Explanation	Resolution
Description Check	⚠️ Warning	The pull request description is empty and does not convey any information about the changeset, so it fails to describe even a part of the modifications introduced.	Please add a brief description summarizing the objectives and key changes of this pull request to help reviewers understand its purpose and scope.
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	You can run @coderabbitai generate docstrings to improve docstring coverage.

✅ Passed checks (1 passed)

Check name	Status	Explanation
Title Check	✅ Passed	The pull request title clearly and concisely highlights the primary change by stating that a custom system prompt parameter is now accepted by the Holmes chat endpoint, and including the ticket identifier adds useful context without unnecessary detail.

✨ Finishing touches

• 📝 Generate Docstrings

🧪 Generate unit tests

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch ROB-2197-allow-passing-custom-sys-prompt-param-in-ask-holmes

---

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]

Actionable comments posted: 0

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

server.py (1)

329-336: Add missing additional_system_prompt parameter to build_chat_messages signature

The build_chat_messages function currently doesn’t accept additional_system_prompt, causing a mismatch when it’s forwarded from server.py. Update its definition (and similarly add_or_update_system_prompt and the ChatRequest model) to include:

def build_chat_messages(..., additional_system_prompt: Optional[str] = None):
    ...

🧹 Nitpick comments (1)

server.py (1)

254-276: Parity: add additional_system_prompt to workload_health_chat

If product scope allows, pass the field into build_workload_health_chat_messages so health checks support the same customization.

📜 Review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between a21e9b3 and 9884890.

📒 Files selected for processing (1)

• server.py (1 hunks)

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (2)

• GitHub Check: Pre-commit checks
• GitHub Check: llm_evals

🔇 Additional comments (1)

server.py (1)

144-153: Extend support to RCA and follow‑ups for parity

Consider threading additional_system_prompt through:

• investigate_issues (/api/investigate)
• stream_investigate_issues (/api/stream/investigate)
• issue_conversation (/api/issue_chat)

This keeps behavior consistent across RCA and follow‑up flows.

Also applies to: 164-189, 285-312

[github-actions]

Results of HolmesGPT evals

• ask_holmes: 34/36 test cases were successful, 1 regressions, 1 setup failures

Test suite	Test case	Status
ask	01_how_many_pods	✅
ask	02_what_is_wrong_with_pod	✅
ask	04_related_k8s_events	✅
ask	05_image_version	✅
ask	09_crashpod	✅
ask	10_image_pull_backoff	✅
ask	110_k8s_events_image_pull	✅
ask	11_init_containers	✅
ask	13a_pending_node_selector_basic	✅
ask	14_pending_resources	✅
ask	15_failed_readiness_probe	✅
ask	17_oom_kill	✅
ask	19_detect_missing_app_details	✅
ask	20_long_log_file_search	✅
ask	24_misconfigured_pvc	✅
ask	24a_misconfigured_pvc_basic	✅
ask	28_permissions_error	🚧
ask	33_cpu_metrics_discovery	❌
ask	39_failed_toolset	✅
ask	41_setup_argo	✅
ask	42_dns_issues_steps_new_tools	✅
ask	43_current_datetime_from_prompt	✅
ask	45_fetch_deployment_logs_simple	✅
ask	51_logs_summarize_errors	✅
ask	53_logs_find_term	✅
ask	54_not_truncated_when_getting_pods	✅
ask	59_label_based_counting	✅
ask	60_count_less_than	✅
ask	61_exact_match_counting	✅
ask	63_fetch_error_logs_no_errors	✅
ask	79_configmap_mount_issue	✅
ask	83_secret_not_found	✅
ask	86_configmap_like_but_secret	✅
ask	93_calling_datadog[0]	✅
ask	93_calling_datadog[1]	✅
ask	93_calling_datadog[2]	✅

Legend

• ✅ the test was successful
• :minus: the test was skipped
• ⚠️ the test failed but is known to be flaky or known to fail
• 🚧 the test had a setup failure (not a code regression)
• 🔧 the test failed due to mock data issues (not a code regression)
• 🚫 the test was throttled by API rate limits/overload
• ❌ the test failed and should be fixed before merging the PR