feat(gemini): add gemini model mapping and always-on triage rules - #7
Merged
Conversation
cristim
added
enhancement
Contributor
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughUpdates CLAUDE.md to add OpenAI and Gemini model-tier mappings, refine subagent delegation to require Changes
Sequence Diagram(s)(Skipped — documentation and workflow changes; no new multi-component runtime control flow to visualize.) Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Review rate limit: 0/5 reviews remaining, refill in 48 minutes and 17 seconds. Comment |
Member
Author
|
@coderabbitai review |
Contributor
✅ Actions performedReview triggered.
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
CLAUDE.md (1)
300-300: ⚡ Quick winPrefer an explicit section reference for the
type/questionexception.Line 300’s “existing rule below” is harder to maintain than a direct reference to the source rule section in
~/.claude/triage.md(as already done intriage.mditself). Tightening this phrasing will reduce cross-file drift.Suggested wording
-- **Exception — `type/question` items** still skip the priority rubric per the existing rule below: apply `type/question` + `status/needs-info`, post the clarifying question, mark `triaged`, and leave open. +- **Exception — `type/question` items** still skip the priority rubric per `~/.claude/triage.md` §"Picking the next thing to work on": apply `type/question` + `status/needs-info`, post the clarifying question, mark `triaged`, and leave open.Based on learnings: Whenever a gap in CLAUDE.md or linked files (coding-standards.md, conventions.md, infra-ops.md, project-docs.md, multi-agent-comms.md, project-level CLAUDE.md) could be improved, proactively propose the change with description and rationale; wait for approval before editing.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@CLAUDE.md` at line 300, Replace the vague phrase "existing rule below" in CLAUDE.md (the line describing the Exception — `type/question` items) with an explicit pointer to the exact section in the triage rules: reference the `~/.claude/triage.md` section that defines the `type/question` exception (mentioning `type/question` + `status/needs-info`) instead of "below" so readers can find the authoritative rule directly; update the sentence to read something like "see the 'type/question' exception in ~/.claude/triage.md (section: EXCEPTION: type/question handling)" to remove ambiguity and prevent cross-file drift.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@CLAUDE.md`:
- Line 223: The text incorrectly refers to "Task tool spawns"; update the
wording to consistently use the Agent tool when describing model
selection—specifically change "Task tool spawns" to "Agent tool spawns" (or
otherwise reference the Agent tool's model parameter) so it aligns with the rule
defined for the Agent tool and its model parameter; ensure the paragraph that
explains matching model tier (Haiku, Sonnet, Opus) explicitly ties those tiers
to the Agent tool's model parameter and includes the guidance "when in doubt, go
one tier cheaper."
---
Nitpick comments:
In `@CLAUDE.md`:
- Line 300: Replace the vague phrase "existing rule below" in CLAUDE.md (the
line describing the Exception — `type/question` items) with an explicit pointer
to the exact section in the triage rules: reference the `~/.claude/triage.md`
section that defines the `type/question` exception (mentioning `type/question` +
`status/needs-info`) instead of "below" so readers can find the authoritative
rule directly; update the sentence to read something like "see the
'type/question' exception in ~/.claude/triage.md (section: EXCEPTION:
type/question handling)" to remove ambiguity and prevent cross-file drift.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
cristim
force-pushed
the
feat/gemini-model-mapping-v2
branch
from
April 30, 2026 20:32
7e4ec9b to
4c486eb
Compare
Member
Author
|
@coderabbitai review |
Contributor
✅ Actions performedReview triggered.
|
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (1)
CLAUDE.md (1)
216-218: ⚡ Quick winInconsistent model identifier format between mapping and usage guidance.
Lines 216-218 use shorthand labels ("OpenAI mini", "OpenAI mid", "OpenAI top") that don't match the specific model identifiers defined in the mapping at lines 19-21 ("gpt-5.4-mini", "gpt-5.4", "gpt-5.5"). Similarly for Gemini, the mapping uses full version numbers.
When an agent reads the tier descriptions to select a model for the
Agenttool'smodelparameter, they need the exact API identifier, not a descriptive shorthand. The mismatch forces cross-referencing back to lines 19-21.♻️ Proposed fix to use consistent model identifiers
-- **Haiku / OpenAI mini / Gemini 3.1 Flash-Lite** (default for most delegations): file renames, typo fixes, mechanical edits with a clear spec, simple lookups (grep for a symbol, find where X is called), reading a single file to answer a factual question, formatting/style fixes, running a single command (or routine `gh`/`git` operations) and reporting output, implementing a tightly-specified function, writing a new test from a tight spec, code review of a small single-file diff, mechanical API/SDK migration where the mapping is documented, classifying or labelling items against a clear rubric (e.g. backlog triage chunks), summarising a single file or short diff. Cheap, fast, good enough when the answer is mostly mechanical or rubric-driven. -- **Sonnet / OpenAI mid / Gemini 3.1 Flash**: focused multi-file changes where coordination across files needs judgement, implementing a function whose spec is mostly clear but has 1–2 design choices, code review of a multi-file diff or a diff with non-trivial logic, refactors with a clear target shape, agentic loops where each step requires non-trivial reasoning (not just rubric-driven labelling). Use when there's some judgement involved but not deep design or hypothesis iteration. -- **Opus / OpenAI top / Gemini 3.1 Pro** (or stay on the current top-level model): architecture decisions, multi-file refactors where the shape is unclear, debugging gnarly bugs that need hypothesis iteration, reading a large unfamiliar codebase from scratch (without `graphify-out/` available) to synthesise a mental model, any work where "understanding" is the hard part rather than the mechanical output. +- **Haiku / gpt-5.4-mini / Gemini 3.1 Flash-Lite** (default for most delegations): file renames, typo fixes, mechanical edits with a clear spec, simple lookups (grep for a symbol, find where X is called), reading a single file to answer a factual question, formatting/style fixes, running a single command (or routine `gh`/`git` operations) and reporting output, implementing a tightly-specified function, writing a new test from a tight spec, code review of a small single-file diff, mechanical API/SDK migration where the mapping is documented, classifying or labelling items against a clear rubric (e.g. backlog triage chunks), summarising a single file or short diff. Cheap, fast, good enough when the answer is mostly mechanical or rubric-driven. +- **Sonnet / gpt-5.4 / Gemini 3.1 Flash**: focused multi-file changes where coordination across files needs judgement, implementing a function whose spec is mostly clear but has 1–2 design choices, code review of a multi-file diff or a diff with non-trivial logic, refactors with a clear target shape, agentic loops where each step requires non-trivial reasoning (not just rubric-driven labelling). Use when there's some judgement involved but not deep design or hypothesis iteration. +- **Opus / gpt-5.5 / Gemini 3.1 Pro** (or stay on the current top-level model): architecture decisions, multi-file refactors where the shape is unclear, debugging gnarly bugs that need hypothesis iteration, reading a large unfamiliar codebase from scratch (without `graphify-out/` available) to synthesise a mental model, any work where "understanding" is the hard part rather than the mechanical output.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@CLAUDE.md` around lines 216 - 218, The tier descriptions use shorthand labels ("OpenAI mini", "OpenAI mid", "OpenAI top", "Gemini 3.1 Flash-Lite") that are inconsistent with the mapping identifiers, so update the descriptive text to use the exact API model IDs from the mapping (e.g., "gpt-5.4-mini", "gpt-5.4", "gpt-5.5" and the full Gemini version strings) wherever the Agent `model` selection guidance is given; specifically edit the paragraph containing the three bullets to replace the shorthand names with the corresponding exact identifiers and keep any human-readable nickname only in parentheses after the exact ID to avoid ambiguity when agents set the Agent tool's model parameter.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@CLAUDE.md`:
- Around line 216-218: The tier descriptions use shorthand labels ("OpenAI
mini", "OpenAI mid", "OpenAI top", "Gemini 3.1 Flash-Lite") that are
inconsistent with the mapping identifiers, so update the descriptive text to use
the exact API model IDs from the mapping (e.g., "gpt-5.4-mini", "gpt-5.4",
"gpt-5.5" and the full Gemini version strings) wherever the Agent `model`
selection guidance is given; specifically edit the paragraph containing the
three bullets to replace the shorthand names with the corresponding exact
identifiers and keep any human-readable nickname only in parentheses after the
exact ID to avoid ambiguity when agents set the Agent tool's model parameter.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: cdc2a96f-33de-45aa-83c4-79383dc2cb20
📒 Files selected for processing (2)
CLAUDE.mdtriage.md
✅ Files skipped from review due to trivial changes (1)
- triage.md
cristim
force-pushed
the
feat/gemini-model-mapping-v2
branch
from
April 30, 2026 20:41
4c486eb to
a7ddc10
Compare
- Add Gemini 1.5 Flash/Pro/Ultra mapping to model tiers - Update subagent strategy to reference OpenAI models - Implement always-on per-item triage rule in CLAUDE.md and triage.md
cristim
force-pushed
the
feat/gemini-model-mapping-v2
branch
from
April 30, 2026 20:42
a7ddc10 to
a90c30e
Compare
Member
Author
|
@coderabbitai review |
Contributor
✅ Actions performedReview triggered.
|
This was referenced May 3, 2026
5 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR adds Gemini model mappings to the model tier rubric and implements the always-on per-item triage rule to improve backlog hygiene.
Summary by CodeRabbit
triagedwhen created, updated, or reviewed; apply the rubric immediately on create/update/read, include full labels on creation, usestatus/needs-info+ a clarifying question when needed, and handletype/questionwith a special flow that marks triaged but leaves the item open.