Heterogeneous Agent Development

Use this skill when the bug or feature lives in the external CLI agent pipeline, not the normal server-side agent runtime.

Use This Skill For

• Adding or changing a driver under apps/desktop/src/main/modules/heterogeneousAgent/drivers/
• Editing an adapter under packages/heterogeneous-agents/src/adapters/
• Changing CLI/Desktop dispatch launchers, including OpenClaw/Hermes, or debugging task-comment authorship
• Debugging heteroAgentRawLine transport, window.__HETERO_AGENT_TRACE, or executeHeterogeneousAgent
• Fixing Claude Code stream-json bugs such as duplicate partial/full chunks, broken message.id boundaries, missing tool_result, TodoWrite state drift, or subagent thread routing
• Fixing Codex JSONL bugs such as mixed multi-tool messages, broken turn boundaries, or missing tool-result mapping
• Fixing step-boundary, tool persistence, subagent thread, or resume bugs in Claude Code / Codex flows
• Reproducing multi-tool mixing, orphan tool messages, or stuck tool-result loading

Pipeline Map

1. CLI raw stdout / JSONL
2. Electron main spawns the CLI and broadcasts heteroAgentRawLine
3. Adapter maps raw provider events into HeterogeneousAgentEvent
4. executeHeterogeneousAgent persists assistant/tool messages and forwards stream events
5. createGatewayEventHandler hydrates the UI
6. Only after this path looks correct should you move on to agent-tracing or context-engine debugging

Read These Files First

• apps/desktop/src/main/controllers/HeterogeneousAgentCtr.ts
• apps/desktop/src/main/modules/heterogeneousAgent/drivers/claudeCode.ts
• apps/desktop/src/main/modules/heterogeneousAgent/drivers/codex.ts
• packages/heterogeneous-agents/src/adapters/claudeCode.ts
• packages/heterogeneous-agents/src/adapters/codex.ts
• src/store/chat/slices/agentRun/actions/transports/hetero/heterogeneousAgentExecutor.ts
• src/store/chat/slices/agentRun/actions/__tests__/heterogeneousAgentExecutor.test.ts

Default Debug Order

1. Prove whether the raw CLI output is correct before touching UI code. The app records every real session — read the most recent one via cat .heerogeneous-tracing/.last-live-trace rather than hand-rolling a claude -p repro (see references/debug-workflow.md §2).
2. If raw output is correct, compare it with adapter output. In dev, executeHeterogeneousAgent exposes window.__HETERO_AGENT_TRACE.
3. If adapted events look correct, inspect persistToolBatch, persistToolResult, step transitions, and subagent routing.
4. Turn the repro into a focused test before fixing.
5. Only after the transport/adapter/executor path looks sound should you debug later-stage message processing.

Critical Invariants

• One raw tool item must map to one stable ToolCallPayload.id.
• A new main-agent step must emit a boundary signal before events are forwarded to the new assistant.
• In Claude Code, multiple assistant events with the same message.id are one turn, not multiple turns.
• In Claude Code, tool_result lives in type: 'user' events, not assistant events.
• In Claude Code partial mode, message_delta.usage is authoritative; do not trust echoed usage on every assistant block.
• persistToolBatch must pre-register assistant tools[] before creating tool messages.
• Every tool message must keep parentId equal to the owning assistant and tool_call_id equal to the tool id.
• tool_result must resolve an existing toolMsgIdByCallId.
• Subagent chunks must stay in thread scope and must not be forwarded into the main assistant stream.
• Never clear the global toolMsgIdByCallId map at main step boundaries.

Dispatched Run Identity

A connector or desktop process can itself run inside another Agent's conversation. Its inherited environment identifies the launcher, not the dispatched Agent.

• Build child identity from the current dispatch. Clear inherited LOBEHUB_AGENT_ID, LOBEHUB_ASSISTANT_MESSAGE_ID, LOBEHUB_TASK_ID, and LOBEHUB_WORKSPACE_ID; replace LOBEHUB_TOPIC_ID and LOBEHUB_OPERATION_ID with the dispatched values. Inject optional context only when supplied by the current dispatch; absent values must not retain ancestor identity.
• Keep authorship independent of authentication. Clearing LOBEHUB_JWT or falling back to user credentials must not change the executing Agent. Never infer the author from the topic owner or task assignee.
• In apps/cli/src/commands/task/lifecycle.ts, lh task comment must refuse to post when topic or operation context exists but the Agent ID is missing or blank. Ordinary terminal comments without execution identity remain user-authored. Misattributed Agent progress can otherwise become human <user_feedback> in later task runs.

Check all four launch surfaces when changing this contract or adding a launcher:

Regression coverage must include a clean environment, conflicting ancestor identity, and missing current Agent/optional context. Check both OpenClaw and Hermes on CLI and Desktop, plus task comments after JWT fallback. Mocked spawn assertions verify environment construction, not end-to-end propagation through an external runtime.

Common Bug Patterns

• Claude Code duplicates text or thinking: check whether partial deltas and the later full assistant block are both being emitted.
• Claude Code opens too many assistant messages: check whether the adapter is cutting steps on every assistant event instead of only on message.id changes.
• Claude Code tool results never land: check whether type: 'user' tool_result blocks are being ignored because the code only inspects assistant events.
• Claude Code TodoWrite cards look stale: check whether synthesized pluginState.todos is being attached at tool-result time.
• Claude Code subagent transcript leaks into the main bubble: check parent_tool_use_id handling and whether subagent chunks are being forwarded to the main gateway handler.
• Multiple Codex tools collapse into one assistant message: first check whether the adapter emits a usable step boundary such as newStep or an equivalent turn-change signal.
• Orphan tool messages: first check step-transition ordering and whether persistToolBatch Phase 1 ran before tool message creation.
• Tool bubble stays loading: look for tool_result for unknown toolCallId and missing result_msg_id backfill.
• Subagent tools show up in the main bubble: check for subagent chunks reaching the main gateway handler.
• Wrong terminal-error guide (e.g. "usage limit reached" shown for a network drop): a classifier is branching on a structured field whose mere presence isn't its meaning. Grep the field across all event states in a real trace before trusting it — see references/debug-workflow.md §8 (CC rate_limit_info rides on status: "allowed" too).

References

• For commands, trace capture, invariants, and focused test commands, read references/debug-workflow.md.
• Live official-model compatibility checks are available only through explicit user invocation of /testing-heterogeneous-agents in Claude Code or $testing-heterogeneous-agents in Codex. Do not automatically load or run that skill during diagnosis. Its failed-cell evidence can be used here for diagnosis or a code fix.