// Interactive Q&A to gather run context before executing a ResearchOps tree node step. Use when a node is about to be run to collect key context the agent will need (e.g. dataset path, design doc, hyperparameters).
| name | node-run-clarify |
| description | Interactive Q&A to gather run context before executing a ResearchOps tree node step. Use when a node is about to be run to collect key context the agent will need (e.g. dataset path, design doc, hyperparameters). |
Gather context from the user through targeted one-at-a-time questions before launching a tree node's run step. The LLM decides what to ask based on the node kind and existing plan details.
POST /api/researchops/projects/:projectId/tree/nodes/:nodeId/run-clarify with:
messages: conversation so far (starts empty []){ done: bool, question: string, options?: string[] }question to user (with options as quick-reply buttons if provided){ role: "user", content: "<answer>" } to messagesdone: trueclarifyMessages is [].POST /api/researchops/projects/:projectId/tree/nodes/:nodeId/run-step with:
clarifyMessages: the full Q&A exchange (or [] if skipped)force, preflightOnly, searchTrialCountThe clarify endpoint adapts questions by node kind:
| Kind | Typical questions |
|---|---|
experiment | Dataset path? Baseline to compare against? Key hyperparameters? |
implementation | Which design doc to follow? Which KB files are relevant? Branch to work on? |
analysis | Which run outputs to analyze? Expected metrics? Comparison baseline? |
knowledge | Which papers/assets to pull in? Specific sections needed? |
search | Search budget? Metric to optimize? Constraint ranges? |
Clarification messages are stored in the run metadata (clarifyContext) so the agent can reference them during execution. They supplement — not replace — the node's existing plan fields.
The ClarificationChat component lives inside VibeNodeWorkbench (always visible for runnable nodes). It resets when the selected node changes. Users can toggle Skip Q&A / Enable Q&A at any time before launching the run.
Autonomous multi-round research review loop. Supports two modes: (1) Plan-driven: takes an implementation plan file, executes TODO items respecting dependency DAG, uses Codex MCP to verify completion of each item. (2) Free-form: iterates review → fix → re-review until positive assessment. Use when user says 'auto review loop', 'review until it passes', or wants iterative improvement.
Show GPU availability across all SSH servers listed in this project's CLAUDE.md. Use when user says "check GPUs", "which GPUs are free", "gpu status", "GPU 状态", or needs to know where to run experiments.
"Register or update an ARIS run on the web dashboard. Internal utility used by other skills and CLAUDE.md auto-registration. Call at skill start to register, at skill end to report completion. If ~/.claude/aris-api.json is absent, does nothing (zero-impact)."
Sync local and remote project files bidirectionally. Handles code, resources, papers, and experiment outputs. Use when user says "sync", "push files", "pull files", "upload code", "download results".
Initialize a new research repository with the recommended AIRS project structure. Use when starting a new project, setting up a repo, or user says "init repo", "scaffold project", "create project structure".
Search and analyze research papers, find related work, summarize key ideas. Use when user says "find papers", "related work", "literature review", "what does this paper say", or needs to understand academic papers.