fix(tools): separate tools job #222
Conversation
WalkthroughThis PR removes MCP server inputs and parsing from agent creation and update paths, narrowing create/update schemas and simplifying createAgentTool to require only an agent profile (promptsId obtained internally). Changes
Sequence Diagram(s)sequenceDiagram
participant UI as User/API
participant Supervisor as createAgentTool
participant Prompts as PromptsService
participant DB as AgentStore
Note over UI,Supervisor: New create flow (MCP fields removed)
UI->>Supervisor: POST createAgent(profile: Profile)
Supervisor->>Prompts: ensurePromptsId(userId)
Prompts-->>Supervisor: promptsId
Supervisor->>DB: createAgent({ profile, prompts_id: promptsId, defaults })
DB-->>Supervisor: agentCreated
Supervisor-->>UI: 201 Created (agent)
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes
Possibly related PRs
Suggested reviewers
Poem
Pre-merge checks and finishing touches❌ Failed checks (1 warning, 1 inconclusive)
✅ Passed checks (1 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
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. Comment |
HxSimo
changed the title
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
packages/agent/src/agents/operators/supervisor/tools/createAgentTool.ts (1)
226-249: Remove the unusedprovidedIdparameter to simplify the function signature.The verification confirms that
ensurePromptsIdis only ever called with a single parameter (userId) at line 108. TheprovidedIdparameter is never passed anywhere in the codebase, making it safe to remove:async function ensurePromptsId(userId: string): Promise<{ id: string; created: boolean }> { const existing = await agents.getExistingPromptsForUser(userId); if (existing) { return { id: existing.id, created: false }; } const promptId = await agents.createDefaultPrompts( userId, TASK_EXECUTOR_SYSTEM_PROMPT, TASK_MANAGER_SYSTEM_PROMPT, TASK_VERIFIER_SYSTEM_PROMPT, TASK_MEMORY_MANAGER_SYSTEM_PROMPT, false ); return { id: promptId, created: true }; }
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
packages/agent/src/agents/operators/supervisor/tools/createAgentTool.ts(3 hunks)packages/agent/src/agents/operators/supervisor/tools/schemas/createAgent.schema.ts(1 hunks)
🧰 Additional context used
🧬 Code graph analysis (2)
packages/agent/src/agents/operators/supervisor/tools/schemas/createAgent.schema.ts (1)
packages/agent/src/agents/operators/supervisor/tools/schemas/common.schemas.ts (1)
AgentProfileSchema(7-32)
packages/agent/src/agents/operators/supervisor/tools/createAgentTool.ts (1)
packages/core/src/common/agent.ts (1)
Input(156-159)
⏰ 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: validate-build
🔇 Additional comments (4)
packages/agent/src/agents/operators/supervisor/tools/createAgentTool.ts (3)
29-31: LGTM! Clear documentation of simplified behavior.The updated description accurately reflects that the tool now only accepts profile configuration and applies defaults for everything else, with a helpful pointer to use
update_agentfor post-creation modifications.
107-113: LGTM! Prompts initialization properly simplified.The prompts are now always derived internally via
ensurePromptsId(userId), which aligns with the PR objective of removing configuration from the input and using defaults.
166-189: ✓ Verification complete—no issues found.
normalizeNumericValuesproperly transformsPartial<AgentConfig.Input>into a completeAgentConfig.Inputwith all required fields (graph,memory,rag,mcp_servers, andprofile) initialized from defaults and then normalized. The return typeNormalizationResult.normalizedConfigis correctly typed asAgentConfig.Input, ensuring type safety. The simplification inbuildAgentConfigFromInputis sound.packages/agent/src/agents/operators/supervisor/tools/schemas/createAgent.schema.ts (1)
5-11: LGTM! Clean, minimal schema with strict validation.The schema is well-designed:
- Reduced to only the required
profilefield, aligning with the PR objective- The
.strict()mode ensures no extraneous fields can be passed, providing clear boundaries- The profile description clearly indicates it's required
| import { AgentProfileSchema } from './common.schemas.js'; | ||
|
|
||
| // Main schema for creating an agent (profile required, other fields optional) | ||
| // Main schema for creating an agent (only profile and prompts_id allowed) |
There was a problem hiding this comment.
Update comment to reflect actual schema.
The comment states "only profile and prompts_id allowed" but the schema only includes the profile field. The prompts_id is now derived internally (see createAgentTool.ts line 113) and is not part of the input schema.
Apply this diff to correct the comment:
-// Main schema for creating an agent (only profile and prompts_id allowed)
+// Main schema for creating an agent (only profile allowed)📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| // Main schema for creating an agent (only profile and prompts_id allowed) | |
| // Main schema for creating an agent (only profile allowed) |
🤖 Prompt for AI Agents
In
packages/agent/src/agents/operators/supervisor/tools/schemas/createAgent.schema.ts
around line 4, update the top comment which incorrectly says "only profile and
prompts_id allowed" to reflect the actual schema (which only includes the
profile field) by removing the mention of prompts_id and/or noting that
prompts_id is derived internally in createAgentTool.ts; modify the comment text
to accurately describe that the input schema accepts only the profile and that
prompts_id is not part of the input.
Summary by CodeRabbit
Refactor
Documentation