Leverage Claude Code subagents

[clay-good]

Note

Moved from issue #216, originally opened by @vialcollet on 2025-10-20. Continuing the conversation here in Discussions — the original issue thread (including any comments) stays available at #216 for the record.

---

Hi all

Has any of you tried to leverage Claude Code subagents with openspec?
This would improve context management as well as increasing accuracy.

I'm on it right now but if you have some experience there I'm highly interested to gather your feedback.

[sandeeppachauri]

🚀 Built a set of open-source accelerators for teams building on the Claude Agent SDK.

After shipping a few Claude-powered agents, I kept solving the same problems from scratch — credential handling, tracing, prompt config, orchestration. So I turned them into reusable, composable building blocks:

🔐 Auth resolution across console API keys, claude login OAuth, and OS-mounted sessions — one call, zero env-var wrangling
📊 Drop-in JSON-line tracing for tool calls & agent activity via hooks — no other dependencies
📝 Prompt-as-config pattern — scope/format/constraints live in YAML, enforced in code, editable without a redeploy
🧩 A master orchestrator with a single execute(payload) entry point, model routing with fallback, and a scaffold CLI to spin up new projects that inherit the whole stack instantly

Everything's independently installable, Python 3.10+, MIT-licensed.

🔗 github.com/sandeeppachauri/Accelerators
🔗 github.com/sandeeppachauri/claude-orchestration-accelerator

this is my attempt at reusable plumbing for the agentic AI era. Feedback and contributions welcome.

#ClaudeAgentSDK #AI #OpenSource #EnterpriseIntegration #SoftwareEngineering

[salgadobreno]

@vialcollet have you tried /subtask, /fork or just making your own agent and routing the task to it? Haven't tried it yet but experienced context-rot on my first run with OpenSpec in a single conversation, now reviewing how subagents work, I'm thinking maybe I could approach managing context differently on my own terms, without changing how OpenSpec is laid out right now because I thought the whole socratic method thing and giving me ample margin for steering was really nice. cc @clay-good