feat(models): add Gemini 3 Flash Preview and improve model search

[Crunchyman-ralph]

• Add gemini-3-flash-preview to Google and Gemini CLI providers
• Use name field from supported-models.json when available
• Improve model search to match both display names and model IDs

What type of PR is this?

• 🐛 Bug fix
• ✨ Feature
• 🔌 Integration
• 📝 Docs
• 🧹 Refactor
• Other:

Description

Related Issues

How to Test This

# Example commands or steps

Expected result:

Contributor Checklist

• Created changeset: npm run changeset
• Tests pass: npm test
• Format check passes: npm run format-check (or npm run format to fix)
• Addressed CodeRabbit comments (if any)
• Linked related issues (if any)
• Manually tested the changes

Changelog Entry

---

For Maintainers

• PR title follows conventional commits
• Target branch correct
• Labels added
• Milestone assigned (if applicable)

Summary by CodeRabbit

• New Features

  • Added support for Gemini 3 Flash Preview model for Google and Gemini CLI providers

• Bug Fixes

  • Improved model search functionality to match both display names and model IDs in the setup command, making it easier to find models

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[changeset-bot]

🦋 Changeset detected

Latest commit: 8067d76

The changes in this PR will be included in the next version bump.

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[coderabbitai]

Walkthrough

The PR adds Gemini 3 Flash Preview model support to the task-master CLI and enhances model search functionality to match both display names and model IDs. It also updates the model data interface and configuration manager to support explicit model names with fallback to ID-based name derivation.

Changes

Cohort / File(s)	Change Summary
Changelog entries \.changeset/floppy-books-feel\.md, \.changeset/frank-ears-remain\.md	Two new changelog entries: one for improved model search matching (patch), another for Gemini 3 Flash Preview support (minor).
Model search filtering apps/cli/src/commands/models/prompts\.ts	Enhanced createSearchSource filter to search by both display name and model ID instead of only choice name, enabling more flexible model discovery.
Model data interface apps/cli/src/lib/model-management\.ts	Added optional name?: string property to ModelData interface.
Model configuration scripts/modules/config-manager\.js	Updated getAvailableModels to prefer explicit name field from JSON when available, falling back to ID-based name generation while preserving special-case overrides.
Supported models catalog scripts/modules/supported-models\.json	Added Gemini 3 Flash Preview (gemini-3-flash-preview) entries to gemini-cli (zero cost, research role allowed) and google (0.5/3.0 cost per 1M tokens) providers with 1048576 max tokens.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Areas requiring extra attention:

• Verify the search logic change in prompts.ts correctly extracts modelId from choice.value structure and handles cases where value may not be an object
• Confirm the fallback behavior in config-manager.js preserves backward compatibility for models without explicit name fields and that override logic is still applied
• Validate Gemini 3 Flash Preview configuration parameters (cost, max_tokens, roles) across both provider entries

Possibly related PRs

• #1427: Adds different Gemini model variants to supported-models.json
• #1317: Modifies supported-models.json and config-manager.js for model definition/loading changes
• #1440: Adds name fields to supported models metadata and updates ModelData interface

Suggested reviewers

• eyaltoledano

Pre-merge checks and finishing touches

✅ 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 summarizes the two main changes: adding Gemini 3 Flash Preview support and improving model search to match both names and IDs.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch ralph/feat/add.gemini.flash

---

📜 Recent review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 38c2c08 and 8067d76.

📒 Files selected for processing (6)

• .changeset/floppy-books-feel.md (1 hunks)
• .changeset/frank-ears-remain.md (1 hunks)
• apps/cli/src/commands/models/prompts.ts (1 hunks)
• apps/cli/src/lib/model-management.ts (1 hunks)
• scripts/modules/config-manager.js (1 hunks)
• scripts/modules/supported-models.json (2 hunks)

🧰 Additional context used

📓 Path-based instructions (12)

**/*.ts

📄 CodeRabbit inference engine (.cursor/rules/test_workflow.mdc)

TypeScript test files must achieve minimum code coverage thresholds: 80% lines/functions and 70% branches globally, 90% for utilities, and 85% for middleware; new features must meet or exceed these thresholds

Files:

• apps/cli/src/lib/model-management.ts
• apps/cli/src/commands/models/prompts.ts

**/*.{js,ts}

📄 CodeRabbit inference engine (.cursor/rules/utilities.mdc)

**/*.{js,ts}: Import and use specific getters from config-manager.js (e.g., getMainProvider(), getLogLevel(), getMainMaxTokens()) to access configuration values needed for application logic
Use isApiKeySet(providerName, session) from config-manager.js to check if a provider's key is available before potentially attempting an AI call
Do not add direct console.log calls outside the logging utility - use the central log function instead
Ensure silent mode is disabled in a finally block to prevent it from staying enabled
Do not access the global silentMode variable directly - use the exported silent mode control functions instead
Do not duplicate task ID formatting logic across modules - centralize formatting utilities
Use ContextGatherer class from utils/contextGatherer.js for AI-powered commands that need project context, supporting tasks, files, custom text, and project tree context
Use FuzzyTaskSearch class from utils/fuzzyTaskSearch.js for automatic task relevance detection with configurable search parameters
Use fuzzy search to supplement user-provided task IDs and display discovered task IDs to users for transparency
Do not replace explicit user task selections with fuzzy results - fuzzy search should supplement, not replace user selections
Use readJSON and writeJSON utilities for all JSON file operations instead of raw fs.readFileSync or fs.writeFileSync
Include error handling for JSON file operations and validate JSON structure after reading
Use path.join() for cross-platform path construction and path.resolve() for absolute paths, validating paths before file operations
Support both .env files and MCP session environment for environment variable resolution with fallbacks for missing values
Prefer updating the core function to accept an outputFormat parameter and check outputFormat === 'json' before displaying UI elements

Files:

• apps/cli/src/lib/model-management.ts
• apps/cli/src/commands/models/prompts.ts
• scripts/modules/config-manager.js

**/*.{ts,tsx}

📄 CodeRabbit inference engine (CLAUDE.md)

Import modules with .js extension even in TypeScript source files for ESM compatibility

Files:

• apps/cli/src/lib/model-management.ts
• apps/cli/src/commands/models/prompts.ts

apps/cli/src/**/*.{ts,tsx}

📄 CodeRabbit inference engine (CLAUDE.md)

CLI (@tm/cli) should be a thin presentation layer that calls tm-core methods and displays results; handle only CLI-specific concerns like argument parsing, output formatting, and user prompts

Files:

• apps/cli/src/lib/model-management.ts
• apps/cli/src/commands/models/prompts.ts

scripts/modules/supported-models.json

📄 CodeRabbit inference engine (.cursor/rules/ai_providers.mdc)

Update scripts/modules/supported-models.json to add a new provider key with an array of model objects, each containing id, name (optional), swe_score, cost_per_1m_tokens (optional), allowed_roles, and max_tokens (optional but recommended)

Files:

• scripts/modules/supported-models.json

scripts/modules/**/*

📄 CodeRabbit inference engine (.cursor/rules/dev_workflow.mdc)

Restart the MCP server if core logic in scripts/modules or MCP tool definitions change

Files:

• scripts/modules/supported-models.json
• scripts/modules/config-manager.js

scripts/modules/config-manager.js

📄 CodeRabbit inference engine (.cursor/rules/ai_providers.mdc)

scripts/modules/config-manager.js: In scripts/modules/config-manager.js, update MODEL_MAP to include the new provider, ensure VALID_PROVIDERS includes the provider, update API key handling in keyMap and the switch statement in getMcpApiKeyStatus and isApiKeySet
For providers not requiring an API key (like Ollama), add a specific check at the beginning of isApiKeySet and getMcpApiKeyStatus to return true immediately for that provider

config-manager.js should load and validate .taskmasterconfig, provide getter functions (getMainProvider, getLogLevel, getDefaultSubtasks, etc.) for accessing settings, manage global.defaultTag and tags section for tag system settings, and NOT directly store or handle API keys

Files:

• scripts/modules/config-manager.js

**/*.js

📄 CodeRabbit inference engine (.cursor/rules/architecture.mdc)

**/*.js: Always use isSilentMode() function to check current silent mode status instead of directly accessing the global silentMode variable or global.silentMode
Use try/finally block pattern when wrapping core function calls with enableSilentMode/disableSilentMode to ensure silent mode is always restored, even if errors occur
For functions that need to handle both a passed silentMode parameter and check global state, check both the function parameter and global state: const isSilent = options.silentMode || (typeof options.silentMode === 'undefined' && isSilentMode())
Functions should accept their dependencies as parameters rather than using globals to promote testability and explicit dependency injection
Define callbacks as separate functions for easier testing rather than inline functions

Files:

• scripts/modules/config-manager.js

scripts/**/*.js

📄 CodeRabbit inference engine (.cursor/rules/context_gathering.mdc)

scripts/**/*.js: Use the ContextGatherer class from scripts/modules/utils/contextGatherer.js to extract context from multiple sources (tasks, files, custom text, project tree) with token counting using gpt-tokens library
Initialize ContextGatherer with project root and tasks path, then call gather() method with tasks array, files array, customContext, includeProjectTree, format ('research', 'chat', or 'system-prompt'), and includeTokenCounts options
Use the FuzzyTaskSearch class from scripts/modules/utils/fuzzyTaskSearch.js for intelligent task discovery with semantic matching, purpose categorization, and relevance scoring using Fuse.js
Implement a three-step initialization pattern for context-aware commands: (1) validate and parse parameters, (2) initialize context gatherer and find project root, (3) auto-discover relevant tasks using fuzzy search if task IDs not specified
Display token breakdown using boxen library with sections for tasks, files, and prompts, showing formatted token counts and file sizes in a clean bordered box with title
Process AI result responses using cli-highlight library to apply syntax highlighting to code blocks with language detection in the format language\ncode
Set reasonable file size limits (50KB default) and project tree depth limits (3-5 levels) when gathering context to maintain performance
Implement graceful error handling for context gathering: handle missing files with warnings, validate task IDs with helpful messages, continue processing if some context sources fail, and provide fallback behavior

Files:

• scripts/modules/config-manager.js

scripts/modules/*.js

📄 CodeRabbit inference engine (.cursor/rules/mcp.mdc)

When implementing MCP support for a command, ensure the core logic function can suppress console output via an outputFormat parameter or other mechanism

scripts/modules/*.js: Use consistent file naming conventions: task_${id.toString().padStart(3, '0')}.txt for task files; use path.join() for composing paths; use appropriate extensions (.txt for tasks, .json for data)
Export all core functions, helper functions, and utility methods needed by dependent code from their respective modules; explicitly verify module export blocks at the bottom of files
Use structured error objects with code and message properties; include clear error messages; handle both function-specific and file system errors; log errors at appropriate severity levels
Use isSilentMode() function to check global silent mode status; wrap core function calls within direct functions using enableSilentMode() and disableSilentMode() in try/finally blocks if the core function produces console output not reliably controlled by outputFormat parameter
Ensure AI calls correctly handle and propagate telemetryData as described in telemetry.mdc
Import context gathering utilities (ContextGatherer, FuzzyTaskSearch) for AI-powered commands; support multiple context types (tasks, files, custom text, project tree); implement detailed token breakdown display
Prefer generateTextService for calls sending large context (like stringified JSON) where incremental display is not needed; import necessary service functions from ai-services-unified.js and prepare parameters (role, session, systemPrompt, prompt)
Create a clear unidirectional flow of dependencies between modules; separate business logic from UI rendering to avoid circular dependencies
Design functions to accept dependencies as parameters; avoid hard-coded dependencies that are difficult to mock
Keep pure logic separate from I/O operations or UI rendering to allow testing logic without mocking complex dependencies
Design core logic to work wi...

Files:

• scripts/modules/config-manager.js

**/*.{js,jsx}

📄 CodeRabbit inference engine (.cursor/rules/test_workflow.mdc)

JavaScript test files using Jest must follow the same testing patterns as TypeScript files, include proper mocking of external dependencies, and achieve the same coverage thresholds

Files:

• scripts/modules/config-manager.js

scripts/modules/**/*.{js,ts}

📄 CodeRabbit inference engine (.cursor/rules/utilities.mdc)

scripts/modules/**/*.{js,ts}: Implement silent migration for tasks.json files that transforms old format to tagged format, marking global flag and performing complete migration
Implement tag resolution functions (getTasksForTag, setTasksForTag, getCurrentTag) that provide backward compatibility with legacy format and default to master tag
Implement complete migration functions for tagged task lists that handle configuration, state file creation, and migration status tracking
When a logger object is passed as a parameter to core functions, ensure the receiving function can call methods like .info, .warn, .error on that object

Files:

• scripts/modules/config-manager.js

🧠 Learnings (29)

📓 Common learnings

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/GEMINI.md:0-0
Timestamp: 2025-11-24T17:56:52.249Z
Learning: Set Gemini 1.5 Flash as fallback model using 'task-master models --set-fallback gemini-1.5-flash' command

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/GEMINI.md:0-0
Timestamp: 2025-11-24T17:56:52.249Z
Learning: Set Gemini as primary model using 'task-master models --set-main gemini-2.0-flash-exp' command

Learnt from: rtmcrc
Repo: eyaltoledano/claude-task-master PR: 933
File: scripts/modules/supported-models.json:238-238
Timestamp: 2025-07-21T14:14:48.694Z
Learning: Model version updates in scripts/modules/supported-models.json may be included in feature PRs if they provide practical improvements like reduced error rates, even if not directly related to the main feature being implemented.

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Applies to assets/**/CLAUDE.md : Update `CLAUDE.md` file with project-specific context, task summaries, and custom command examples for Claude Code integration

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/glossary.mdc:0-0
Timestamp: 2025-11-24T18:00:32.617Z
Learning: Refer to new_features.mdc for guidelines on integrating new features into the Task Master CLI with tagged system considerations

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/taskmaster.mdc:0-0
Timestamp: 2025-11-24T18:02:22.305Z
Learning: Applies to .taskmaster/config.json : The .taskmaster/config.json file stores AI model configuration (main, research, fallback models) managed via 'task-master models' command

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Applies to assets/.taskmaster/config.json : Store AI model configuration in `.taskmaster/config.json` and use `task-master models` command to modify instead of manual editing

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_providers.mdc:0-0
Timestamp: 2025-11-24T17:57:14.743Z
Learning: Use the `models` MCP tool or the `task-master models` CLI command to manage AI configurations

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Applies to assets/.taskmaster/docs/prd.md : Use `task-master parse-prd .taskmaster/docs/prd.md` to generate tasks from PRD document

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Use `gh` CLI to create PRs referencing Task Master task IDs in titles and descriptions following pattern 'Complete task <id>: <description>'

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Use `--research` flag with Task Master commands like `parse-prd`, `add-task`, `expand`, and `update-task` to enhance AI-powered operations with research-based information

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/GEMINI.md:0-0
Timestamp: 2025-11-24T17:56:52.249Z
Learning: Applies to assets/.gemini/settings.json : Configure Task Master MCP server in ~/.gemini/settings.json with the command 'npx' and args ['-y', 'task-master-ai']

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1178
File: packages/tm-core/src/auth/config.ts:5-7
Timestamp: 2025-09-02T21:51:27.921Z
Learning: The user Crunchyman-ralph prefers not to use node: scheme imports (e.g., 'node:os', 'node:path') for Node.js core modules and considers suggestions to change bare imports to node: scheme as too nitpicky.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1069
File: .changeset/fix-tag-complexity-detection.md:0-0
Timestamp: 2025-08-02T15:33:22.656Z
Learning: For changeset files (.changeset/*.md), Crunchyman-ralph prefers to ignore formatting nitpicks about blank lines between frontmatter and descriptions, as he doesn't mind having them and wants to avoid such comments in future reviews.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1132
File: .github/workflows/weekly-metrics-discord.yml:81-93
Timestamp: 2025-08-13T22:10:46.958Z
Learning: Crunchyman-ralph ignores YAML formatting nitpicks about trailing spaces when there's no project-specific YAML formatter configured, preferring to focus on functionality over cosmetic formatting issues.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1132
File: .github/workflows/weekly-metrics-discord.yml:81-93
Timestamp: 2025-08-13T22:10:46.958Z
Learning: Crunchyman-ralph ignores YAML formatting nitpicks about trailing spaces when there's no project-specific YAML formatter configured, preferring to focus on functionality over cosmetic formatting issues.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1105
File: scripts/modules/supported-models.json:242-254
Timestamp: 2025-08-08T11:33:15.297Z
Learning: Preference: In scripts/modules/supported-models.json, the "name" field is optional. For OpenAI entries (e.g., "gpt-5"), Crunchyman-ralph prefers omitting "name" when the id is explicit enough; avoid nitpicks requesting a "name" in such cases.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1200
File: src/ai-providers/custom-sdk/grok-cli/language-model.js:96-100
Timestamp: 2025-09-19T16:06:42.182Z
Learning: The user Crunchyman-ralph prefers to keep environment variable names explicit (like GROK_CLI_API_KEY) rather than supporting multiple aliases, to avoid overlap and ensure clear separation between different CLI implementations.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1178
File: packages/tm-core/src/subpath-exports.test.ts:6-9
Timestamp: 2025-09-03T12:45:30.724Z
Learning: The user Crunchyman-ralph prefers to avoid overly nitpicky or detailed suggestions in code reviews, especially for test coverage of minor import paths. Focus on more substantial issues rather than comprehensive coverage of all possible edge cases.

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1217
File: apps/cli/src/index.ts:16-21
Timestamp: 2025-09-18T16:35:35.147Z
Learning: The user Crunchyman-ralph considers suggestions to export types for better ergonomics (like exporting UpdateInfo type alongside related functions) as nitpicky and prefers not to implement such suggestions.

📚 Learning: 2025-11-24T17:57:14.743Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_providers.mdc:0-0
Timestamp: 2025-11-24T17:57:14.743Z
Learning: Applies to scripts/modules/supported-models.json : Update `scripts/modules/supported-models.json` to add a new provider key with an array of model objects, each containing `id`, `name` (optional), `swe_score`, `cost_per_1m_tokens` (optional), `allowed_roles`, and `max_tokens` (optional but recommended)

Applied to files:

• apps/cli/src/lib/model-management.ts
• scripts/modules/supported-models.json
• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T18:02:49.782Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/telemetry.mdc:0-0
Timestamp: 2025-11-24T18:02:49.782Z
Learning: The telemetryData object structure must include: timestamp (ISO string), userId, commandName, modelUsed, providerName, inputTokens, outputTokens, totalTokens, totalCost, and currency fields

Applied to files:

• apps/cli/src/lib/model-management.ts

📚 Learning: 2025-08-08T11:34:45.482Z

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1105
File: .changeset/vast-weeks-fetch.md:5-5
Timestamp: 2025-08-08T11:34:45.482Z
Learning: In this repo, supported-models.md is auto-generated by CI from supported-models.json; do not request manual edits to that file—ensure JSON entries are correct instead.

Applied to files:

• scripts/modules/supported-models.json
• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T17:56:52.249Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/GEMINI.md:0-0
Timestamp: 2025-11-24T17:56:52.249Z
Learning: Set Gemini 1.5 Flash as fallback model using 'task-master models --set-fallback gemini-1.5-flash' command

Applied to files:

• scripts/modules/supported-models.json
• .changeset/frank-ears-remain.md

📚 Learning: 2025-07-21T14:14:48.694Z

Learnt from: rtmcrc
Repo: eyaltoledano/claude-task-master PR: 933
File: scripts/modules/supported-models.json:238-238
Timestamp: 2025-07-21T14:14:48.694Z
Learning: Model version updates in scripts/modules/supported-models.json may be included in feature PRs if they provide practical improvements like reduced error rates, even if not directly related to the main feature being implemented.

Applied to files:

• scripts/modules/supported-models.json
• .changeset/floppy-books-feel.md
• scripts/modules/config-manager.js

📚 Learning: 2025-08-08T11:34:45.482Z

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1105
File: .changeset/vast-weeks-fetch.md:5-5
Timestamp: 2025-08-08T11:34:45.482Z
Learning: In this repo, the supported models list is auto-generated by CI into docs/models.md from scripts/modules/supported-models.json via .github/workflows/update-models-md.yml and docs/scripts/models-json-to-markdown.js. Don’t request manual edits to the Markdown; ensure the JSON is correct instead.

Applied to files:

• scripts/modules/supported-models.json
• scripts/modules/config-manager.js

📚 Learning: 2025-08-08T11:34:45.482Z

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1105
File: .changeset/vast-weeks-fetch.md:5-5
Timestamp: 2025-08-08T11:34:45.482Z
Learning: This repo’s supported models list is CI-generated into docs/models.md from scripts/modules/supported-models.json (workflow: .github/workflows/update-models-md.yml using docs/scripts/models-json-to-markdown.js). Don’t request manual edits to the Markdown; ensure the JSON is correct instead. README links to docs/models.md.

Applied to files:

• scripts/modules/supported-models.json
• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T17:56:52.249Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/GEMINI.md:0-0
Timestamp: 2025-11-24T17:56:52.249Z
Learning: Set Gemini as primary model using 'task-master models --set-main gemini-2.0-flash-exp' command

Applied to files:

• scripts/modules/supported-models.json
• .changeset/frank-ears-remain.md

📚 Learning: 2025-12-11T21:01:41.112Z

Learnt from: ben-vargas
Repo: eyaltoledano/claude-task-master PR: 1508
File: scripts/modules/supported-models.json:154-189
Timestamp: 2025-12-11T21:01:41.112Z
Learning: For codex-cli provider in scripts/modules/supported-models.json, the "minimal" reasoning_effort value is NOT supported for gpt-5.1-codex-max and gpt-5.2 models, despite what OpenAI API documentation may indicate. The correct reasoning_efforts array for these models is ["none", "low", "medium", "high", "xhigh"].

Applied to files:

• scripts/modules/supported-models.json

📚 Learning: 2025-11-24T17:58:19.853Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/changeset.mdc:0-0
Timestamp: 2025-11-24T17:58:19.853Z
Learning: Provide a concise, single-line changeset summary in imperative mood (e.g., 'Add feature X', 'Fix bug Y') that describes what changed from a user/consumer perspective, not implementation details

Applied to files:

• .changeset/frank-ears-remain.md
• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T17:56:52.249Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/GEMINI.md:0-0
Timestamp: 2025-11-24T17:56:52.249Z
Learning: Applies to assets/.gemini/settings.json : Configure Task Master MCP server in ~/.gemini/settings.json with the command 'npx' and args ['-y', 'task-master-ai']

Applied to files:

• .changeset/frank-ears-remain.md

📚 Learning: 2025-12-11T14:45:14.973Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Applies to assets/.taskmaster/config.json : Store AI model configuration in `.taskmaster/config.json` and use `task-master models` command to modify instead of manual editing

Applied to files:

• .changeset/frank-ears-remain.md
• .changeset/floppy-books-feel.md
• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T18:02:22.305Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/taskmaster.mdc:0-0
Timestamp: 2025-11-24T18:02:22.305Z
Learning: Applies to .taskmaster/config.json : The .taskmaster/config.json file stores AI model configuration (main, research, fallback models) managed via 'task-master models' command

Applied to files:

• .changeset/frank-ears-remain.md
• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T17:57:14.743Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_providers.mdc:0-0
Timestamp: 2025-11-24T17:57:14.743Z
Learning: Task Master uses three roles for AI models: `main` (primary model for general tasks), `research` (model with web access when --research flag is used), and `fallback` (model used if primary fails). Each role is configured with a specific `provider:modelId` pair.

Applied to files:

• .changeset/frank-ears-remain.md
• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T17:57:14.743Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_providers.mdc:0-0
Timestamp: 2025-11-24T17:57:14.743Z
Learning: Use the `models` MCP tool or the `task-master models` CLI command to manage AI configurations

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T18:04:43.972Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/utilities.mdc:0-0
Timestamp: 2025-11-24T18:04:43.972Z
Learning: Applies to **/*.{js,ts} : Use fuzzy search to supplement user-provided task IDs and display discovered task IDs to users for transparency

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T18:00:06.827Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/dev_workflow.mdc:0-0
Timestamp: 2025-11-24T18:00:06.827Z
Learning: Update dependent tasks when implementation differs from original plan using `task-master update --from=<id>` or `task-master update-task --id=<id>`

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T18:00:06.827Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/dev_workflow.mdc:0-0
Timestamp: 2025-11-24T18:00:06.827Z
Learning: Applies to .taskmaster/config.json : Store non-API key settings like model selections and `MAX_TOKENS` in `.taskmaster/config.json`, managed via `task-master models` command, not environment variables

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-12-11T14:45:14.973Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/AGENTS.md:0-0
Timestamp: 2025-12-11T14:45:14.973Z
Learning: Use `--research` flag with Task Master commands like `parse-prd`, `add-task`, `expand`, and `update-task` to enhance AI-powered operations with research-based information

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T18:00:06.827Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/dev_workflow.mdc:0-0
Timestamp: 2025-11-24T18:00:06.827Z
Learning: When implementation differs significantly from planned approach, use `task-master update --from=<futureTaskId> --research` to update multiple future tasks

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-11-24T18:05:02.114Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: assets/.windsurfrules:0-0
Timestamp: 2025-11-24T18:05:02.114Z
Learning: Update dependent tasks when implementation differs from original plan using `task-master update`

Applied to files:

• .changeset/floppy-books-feel.md

📚 Learning: 2025-08-08T11:33:15.297Z

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1105
File: scripts/modules/supported-models.json:242-254
Timestamp: 2025-08-08T11:33:15.297Z
Learning: Preference: In scripts/modules/supported-models.json, the "name" field is optional. For OpenAI entries (e.g., "gpt-5"), Crunchyman-ralph prefers omitting "name" when the id is explicit enough; avoid nitpicks requesting a "name" in such cases.

Applied to files:

• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T17:57:31.417Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_services.mdc:0-0
Timestamp: 2025-11-24T17:57:31.417Z
Learning: Applies to scripts/modules/task-manager/*.js, scripts/modules/commands.js : Do not fetch AI-specific parameters (model ID, max tokens, temperature) using `config-manager.js` getters for AI calls; pass the `role` parameter instead to the unified service

Applied to files:

• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T17:57:14.743Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_providers.mdc:0-0
Timestamp: 2025-11-24T17:57:14.743Z
Learning: Applies to scripts/modules/config-manager.js : In `scripts/modules/config-manager.js`, update `MODEL_MAP` to include the new provider, ensure `VALID_PROVIDERS` includes the provider, update API key handling in `keyMap` and the `switch` statement in `getMcpApiKeyStatus` and `isApiKeySet`

Applied to files:

• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T18:04:43.972Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/utilities.mdc:0-0
Timestamp: 2025-11-24T18:04:43.972Z
Learning: Applies to scripts/modules/task-manager/**/*.{js,ts} : Do not call AI-specific getters (like getMainModelId, getMainMaxTokens) from core logic functions in scripts/modules/task-manager/*. Instead, pass the role to the unified AI service

Applied to files:

• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T17:57:14.743Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_providers.mdc:0-0
Timestamp: 2025-11-24T17:57:14.743Z
Learning: When setting a model ID using the `models` MCP tool or CLI command, DO NOT include the `provider:` prefix—use only the model ID itself (e.g., `gpt-4o` not `openai:gpt-4o`)

Applied to files:

• scripts/modules/config-manager.js

📚 Learning: 2025-10-01T19:53:34.261Z

Learnt from: Crunchyman-ralph
Repo: eyaltoledano/claude-task-master PR: 1262
File: scripts/modules/task-manager/update-tasks.js:216-233
Timestamp: 2025-10-01T19:53:34.261Z
Learning: For scripts/modules/task-manager/*.js: Use generateObjectService with Zod schemas for structured AI responses rather than generateTextService + manual JSON parsing, as modern AI providers increasingly support the tool use and generateObject paradigm with improved reliability.

Applied to files:

• scripts/modules/config-manager.js

📚 Learning: 2025-11-24T17:57:31.417Z

Learnt from: CR
Repo: eyaltoledano/claude-task-master PR: 0
File: .cursor/rules/ai_services.mdc:0-0
Timestamp: 2025-11-24T17:57:31.417Z
Learning: Applies to scripts/modules/task-manager/*.js, scripts/modules/commands.js : Use `config-manager.js` getters only for non-AI related application logic (e.g., `getDefaultSubtasks`), not for fetching AI parameters; the unified service handles AI parameter resolution internally based on the `role`

Applied to files:

• scripts/modules/config-manager.js

🧬 Code graph analysis (1)

apps/cli/src/commands/models/prompts.ts (1)

apps/cli/src/commands/models/types.ts (1)

• ModelChoice (69-74)

⏰ 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). (1)

• GitHub Check: Test

🔇 Additional comments (6)

.changeset/frank-ears-remain.md (1)

1-5: LGTM!

The changeset follows proper conventions with a minor version bump and clear description of the Gemini 3 Flash Preview model addition.

.changeset/floppy-books-feel.md (1)

1-5: LGTM!

The changeset correctly uses a patch version bump for the search improvement and provides a clear description of the enhancement.

apps/cli/src/lib/model-management.ts (1)

21-21: LGTM!

The addition of the optional name field to the ModelData interface is clean and backward-compatible, supporting explicit model naming while preserving existing ID-based name generation as a fallback.

apps/cli/src/commands/models/prompts.ts (1)

161-169: LGTM!

The enhanced search filtering logic correctly combines both display name and model ID for matching, improving the user experience when searching for models. The type checking is safe and handles cases where the modelId might not be present.

scripts/modules/config-manager.js (1)

1052-1068: LGTM!

The naming logic enhancement correctly prefers explicit name fields from the model JSON when available, while maintaining backward-compatible fallback to ID-based name generation with existing special-case overrides. This supports better model display names in the CLI.

scripts/modules/supported-models.json (1)

205-216: Model specifications are accurate and aligned with official Google documentation. No corrections needed.

---

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