Amazing to see your workflow and automation you have set up with CC. Wanted to mention that MCPs are coming to Traycer soon

That will be a dream - I saw the announcement

has multi-agent single-agent been removed?

Yes, we removed support last week. The recommended way for larger tasks is to use Phase breakdown. For complex tasks, we'll soon be adding support for premium models, e.g. you would be able to plan with Opus-4.1 + o3-pro That said, the default planner already uses sub-agents where it makes sense

What does the Traycer option do in the dropdown when selecting your IDE? Does Traycer have its own implementation flow?

Can you use CC in Ci/CD I wonder? I don't know, feels like CC is overhyped. Hooks are pretty unique but subagents are just prompts that run in their own context. I'm going to try it this week I think for its ability to do hooks so I can type check and lint when each Traycer subtask is done.

Yes, traycer can do code generation. It's great for simple tasks, but is missing few capabities that are standard in other coding agents like automatic lint fixes and ability to run terminal commands

Curious how Traycer can be integrated into this new archon system that just came out. Seems like a gamechanger for task management.

Where is the announcement about Traycer getting MCP support?

@brave-gpu-62353, we are actively working on supporting MCPs in Traycer. ETA 7-10 days. We'll make an announcement here

Great. I'm waiting for the re-verify/review stuff more though right now; feel like I'm going in a loop so what I've been doing now is just verifying, implementing the fixes that make sense it suggests and moving on; no matter how many times I re-verify it's never complete.

I am planning to finally try traycer out. Currently my set up has been using cursor + claude code cli inside the cursor. I use cursor ask mode to generate the task plan and then copy/paste that plan into claude code cli for execution. So fsr it has worked decently. But traycer has intrigued me lately, could someone share what is your workflow with cursor + claude cli+traycer? Would love to hear usecases.

@billowy-terabyte-89587 Good day... If your team currently looking to boost your presence through social media marketing, influencer campaigns, community growth (Telegram/Discord/Binance / X), or app download app register Kol service etc. I'd love to collaborate.

I'm also using the same tools, cursor + claude code + traycer. Traycer - I give my feature query to it, and it breaks down into phases; I then chat about missing points, and each phase has a plan. After that, I hand off to Claude code (dont need to copy/paste, its a simple click) Claude code - It follows the plan properly and implements it. Then, Traycer verifies the changes again and I move to the next phase. Cursor - Mainly used for quick changes. Once everything is done by Claude code and Traycer, I fix minor issues with the help of autocomplete and in-line chat (cmd + k).

i used traycer to plan and execute with kilo code. Then at some point, traycer stucked at "planning" for a very long time, so i closed the vs code. is there any way to retrieve the previous "plan" / conversation so that traycer knows the context? And its halfway through the execution, so if i asked traycer to do anything, lots of things will need to "relearn", is that right? How to overcome this?

for some reason, tracyer is not loading in kilocode. what is this? https://cdn.discordapp.com/attachments/1230383226376163331/1409188085492879462/image.png?ex=68ac782d&is=68ab26ad&hm=b7056b899a81196d3dc9cb2ec60d53c8ae068bb94a257a119d271feff9c319a3&

nvm, its actually located at the top right, the history button, able to see the previous task.

You would find the existing tasks in history. We'll gather more details for the other issue https://cdn.discordapp.com/attachments/1230383226376163331/1409229762584383538/history.png?ex=68ac9efd&is=68ab4d7d&hm=cfeb36a177415adf5338457f13af414b137a8141554d05faa9245c67e5fcd431&

i still dont really get the flow

so u plan --> make changes to plan until you like the plan

then u want to plan each segment from the plan doesnt seem to work for me at the moment when i try segment it doesnt generate inside each segment

Hey, we have 2 major workflows. 1. Plan mode - you simply start with a task query and then Traycer explores the codebase, generating a file-level detailed plan with proper descriptions, references, mermaid diagrams, reasoning, etc. You can then iterate on the plan or chat with it to ask anything. Once your plan is final, you can execute it using any of your favorite coding agents like Cursor, Claude Code, Github Copilot, and others. After execution, these agents will follow the plan properly. You can then return to Traycer and click verify changes; Traycer will check if all modifications made by agents adhere to the plan. If any items are missing, Traycer suggests changes directly on the UI in the form of comments. You can use these comments to hand off the fixes to your agent. 2. Phase mode - this is for a step-by-step approach. You add a query, and Traycer understands your intent and asks questions if needed. It then creates multiple phases. You generate a plan for each phase, similar to Plan Mode, where you get everything. After verifying a phase, you move on to the next, keeping the context of previous phases. Here's a quick video by a developer:

Good video. Re-verifying seems to loop a lot, suggesting fixes that a previous verify flagged. Actually, for clarification, if Traycer thinks there are no changes to be made, re-verifying shouldn't be available, making it clear that it's okay to move to the next phase.

We are currently working on an improved version of the re-verification process, which will be released soon.

guys can you check augment prompt enhacer its works also so well uses your codebase for prompt enhancng, many ide are introducing this so will not look for other tool outside their ide so how will you fight with this

https://www.augmentcode.com/blog/prompt-enhancer-live-in-augment-chat https://lukew.com/ff/entry.asp?2101

I just tried it, and my impression is that it's a shallow prompt re-writing feature using a one-shot RAG mechanism.

● You are a prompt enhancement specialist with deep expertise in AI instruction optimization. Your primary function is to transform unclear, ambiguous, or incomplete prompts into precise, actionable, and effective instructions. Core Responsibilities: 1. Critical Analysis - Systematically identify and catalog specific issues including: - Ambiguous terminology or vague language - Missing essential context or background information - Structural problems or logical inconsistencies - Unclear success criteria or expected outcomes 2. Strategic Rewriting - Reconstruct prompts with: - Precise, unambiguous language - Clear action verbs and specific directives - Logical flow and coherent structure - Measurable objectives and deliverables 3. Comprehensive Enhancement - Add critical missing elements: - Detailed context and background information - Specific constraints, limitations, and requirements - Clear formatting and output specifications - Relevant examples or templates when beneficial - Edge case considerations and error handling 4. Intent Preservation - Maintain the original prompt's core purpose while dramatically improving its effectiveness and reducing misinterpretation risk 5. Professional Structuring - Organize enhanced prompts using: - Clear hierarchical sections with descriptive headers - Numbered steps for sequential processes - Bullet points for lists and requirements - Code blocks for technical specifications - Bold/italic formatting for emphasis

Enhancement Methodology: - Audience Analysis - Consider the target user's expertise level, domain knowledge, and technical capabilities - Output Specification - Define exact deliverables, formats, and quality standards - Constraint Definition - Clearly articulate boundaries, limitations, and requirements - Context Integration - Incorporate necessary background information and situational factors - Example Integration - Provide concrete examples or templates to clarify expectations - Scenario Planning - Address potential edge cases, exceptions, and alternative pathways Quality Standards: - Every instruction must be actionable and specific - Eliminate all ambiguous language and undefined terms - Ensure logical consistency throughout the prompt - Provide clear success criteria and evaluation methods - Maintain professional tone while being comprehensive Always preserve the fundamental objective of the original prompt while transforming it into a superior, more effective instruction that minimizes confusion and maximizes successful outcomes.

its useful in some cases