chore(agents): align model names with Cursor subagent configuration by deantzzz · Pull Request #249 · affaan-m/everything-claude-code

deantzzz · 2026-02-18T20:12:29Z

Update architect, build-error-resolver, code-reviewer, database-reviewer
Update doc-updater, e2e-runner, go-build-resolver, go-reviewer
Update planner, python-reviewer, refactor-cleaner, security-reviewer, tdd-guide
Align with Cursor subagent model configuration

Description

Type of Change

Checklist

Tests pass locally (node tests/run-all.js)
Validation scripts pass
Follows conventional commits format
Updated relevant documentation

Summary by CodeRabbit

Chores
- Updated AI agent configurations to newer Claude model variants for improved capability.
- Standardized and expanded tool access lists across agents (Read/Write/Edit/Bash/Grep/Glob as applicable).
- Enhanced agent descriptions to emphasize proactive guidance, focused workflows, and clearer usage expectations.

coderabbitai · 2026-02-18T20:14:03Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

Configuration used: defaults

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 4c5973e and 6cae108.

📒 Files selected for processing (13)

.cursor/agents/architect.md
.cursor/agents/build-error-resolver.md
.cursor/agents/code-reviewer.md
.cursor/agents/database-reviewer.md
.cursor/agents/doc-updater.md
.cursor/agents/e2e-runner.md
.cursor/agents/go-build-resolver.md
.cursor/agents/go-reviewer.md
.cursor/agents/planner.md
.cursor/agents/python-reviewer.md
.cursor/agents/refactor-cleaner.md
.cursor/agents/security-reviewer.md
.cursor/agents/tdd-guide.md

🚧 Files skipped from review as they are similar to previous changes (12)

.cursor/agents/go-reviewer.md
.cursor/agents/tdd-guide.md
.cursor/agents/e2e-runner.md
.cursor/agents/go-build-resolver.md
.cursor/agents/build-error-resolver.md
.cursor/agents/refactor-cleaner.md
.cursor/agents/planner.md
.cursor/agents/python-reviewer.md
.cursor/agents/code-reviewer.md
.cursor/agents/architect.md
.cursor/agents/doc-updater.md
.cursor/agents/security-reviewer.md

📝 Walkthrough

Walkthrough

Standardized and updated YAML front matter across multiple .cursor/agents/*.md files: replaced legacy model identifiers with Claude variants, added/standardized tools arrays, and expanded agent description text; no runtime code or behavioral logic changed.

Changes

Cohort / File(s)	Summary
Architect & Planner `.cursor/agents/architect.md`, `.cursor/agents/planner.md`	Replaced model with `claude-4.6-opus-high-thinking`, added/standardized `tools` arrays, expanded description text; content body unchanged.
Build / Resolver / Runner `.cursor/agents/build-error-resolver.md`, `.cursor/agents/go-build-resolver.md`, `.cursor/agents/e2e-runner.md`	Updated model entries to Claude variants (`claude-4.6-opus-high` / sonnet-thinking), added comprehensive `tools` lists, updated descriptions; metadata-only edits.
Reviewers & Cleaners `.cursor/agents/code-reviewer.md`, `.cursor/agents/python-reviewer.md`, `.cursor/agents/go-reviewer.md`, `.cursor/agents/security-reviewer.md`, `.cursor/agents/refactor-cleaner.md`	Switched model to `claude-4.6-sonnet-medium` (where applicable), moved/standardized `tools` fields and adjusted descriptions; header metadata changes only.
Specialized Agents `.cursor/agents/database-reviewer.md`, `.cursor/agents/doc-updater.md`, `.cursor/agents/tdd-guide.md`	Added/standardized `tools`, updated `model` (sonnet/haiku variants) and expanded domain-specific descriptions (DB, docs, TDD); no functional code changes.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

fix: resolve multiple reported issues (#205, #182, #188, #172, #173) #207 — Modifies the same agent front-matter fields (tools/model) across .cursor/agents/*.md.
feat: add comprehensive Go development support with agents, commands, and skills #83 — Touches the same Go agent files (go-reviewer.md, go-build-resolver.md) and introduces related agent content.
feat: add missing CI/CD pipeline files #97 — Adds/validates agent front-matter schema used by these metadata updates (validator/schema changes).

Suggested labels

needs-fix

Suggested reviewers

affaan-m

Poem

🐰 I hopped through headers, tidy and quick,
Models swapped gently, tools added in a flick.
Descriptions now kinder, metadata bright—
I nibble at diffs, then hop off into night. ✨

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main change: updating agent configuration files to align model names with Cursor subagent configuration standards across all modified agent definitions.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

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

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Post copyable unit tests in a comment

Tip

Issue Planner is now in beta. Read the docs and try it out! Share your feedback on Discord.

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

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

coderabbitai

Actionable comments posted: 2

🧹 Nitpick comments (2)

.cursor/agents/build-error-resolver.md (1)
4-4: Opus-tier model is disproportionate for a minimal-diff build fix agent

Regardless of model name validity (flagged above), the upgrade from sonnet to an Opus-tier model for this agent is worth scrutinising. The agent's own mandate is "smallest possible changes," "no architecture changes," and "fix one error at a time" — mechanical tasks where Sonnet-class models perform equivalently at significantly lower cost. Opus is reserved for open-ended reasoning (architect, planner). If the intent is to ensure the latest Sonnet 4.6 is used, the correct alias would be claude-sonnet-4-6 or sonnet.
✏️ Suggested change
-model: claude-4.6-opus-high
+model: claude-sonnet-4-6
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.

In @.cursor/agents/build-error-resolver.md at line 4, The model entry currently
uses an Opus-tier value ("model: claude-4.6-opus-high") which is
over-provisioned for this minimal-diff build-fix agent; replace that value with
a Sonnet-class alias such as "claude-sonnet-4-6" (or "sonnet" if your system
expects that shorthand) by editing the model key in the file so the agent runs
on a Sonnet-tier model instead of an Opus-tier one.
.cursor/agents/e2e-runner.md (1)
4-4: -thinking (extended thinking) is a mismatch for an execution-oriented E2E agent

Extended thinking is appropriate for deep-reasoning roles (architect, planner). The e2e-runner is an execution specialist — it invokes CLI tools, captures artifacts, and checks test results — tasks where extended thinking adds latency and cost without a corresponding quality benefit. Among the peer agents, only the two reasoning-heavy ones (architect, planner) use a -thinking variant.
✏️ Suggested change
-model: claude-4.6-sonnet-medium-thinking
+model: claude-4.6-sonnet-medium
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.

In @.cursor/agents/e2e-runner.md at line 4, The model string in
.cursor/agents/e2e-runner.md uses the extended-thinking variant
"claude-4.6-sonnet-medium-thinking" which is inappropriate for an
execution-oriented E2E agent; update the model entry to the non-thinking
execution variant by replacing "claude-4.6-sonnet-medium-thinking" with
"claude-4.6-sonnet-medium" so the e2e-runner uses the lower-latency,
non-extended thinking model.

🤖 Prompt for all review comments with AI agents

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

Inline comments:
In @.cursor/agents/refactor-cleaner.md:
- Line 4: Replace the non-standard model identifiers in
.cursor/agents/refactor-cleaner.md (claude-4.6-sonnet-medium,
claude-4.6-opus-high, claude-4.6-opus-high-thinking,
claude-4.6-sonnet-medium-thinking, claude-4.5-haiku) with
Cursor/Anthropic-compliant names (e.g., claude-sonnet-4-6, claude-opus-4-6 or
Cursor registry names like claude-4-sonnet-thinking / claude-4-opus-thinking) OR
add a comment/link to the Cursor models registry (GET
https://api.cursor.com/v0/models) confirming these exact identifiers are valid;
update every agent entry that references those strings to the corrected
identifiers or include the verification link next to the model field so all
agents use validated model identifiers.

In @.cursor/agents/tdd-guide.md:
- Line 2: The tools array for the TDD guide agent is missing the "Glob" tool
which is required for glob-pattern file discovery; update the tools declaration
(the tools: [...] entry in .cursor/agents/tdd-guide.md for the TDD guide agent)
to include "Glob" alongside "Read", "Write", "Edit", "Bash", and "Grep" so the
agent can locate test files like **/*.test.ts and **/*.spec.ts and match the
other write-capable agents.

---

Nitpick comments:
In @.cursor/agents/build-error-resolver.md:
- Line 4: The model entry currently uses an Opus-tier value ("model:
claude-4.6-opus-high") which is over-provisioned for this minimal-diff build-fix
agent; replace that value with a Sonnet-class alias such as "claude-sonnet-4-6"
(or "sonnet" if your system expects that shorthand) by editing the model key in
the file so the agent runs on a Sonnet-tier model instead of an Opus-tier one.

In @.cursor/agents/e2e-runner.md:
- Line 4: The model string in .cursor/agents/e2e-runner.md uses the
extended-thinking variant "claude-4.6-sonnet-medium-thinking" which is
inappropriate for an execution-oriented E2E agent; update the model entry to the
non-thinking execution variant by replacing "claude-4.6-sonnet-medium-thinking"
with "claude-4.6-sonnet-medium" so the e2e-runner uses the lower-latency,
non-extended thinking model.

.cursor/agents/refactor-cleaner.md

.cursor/agents/tdd-guide.md

affaan-m