Skip to main content
prompt GP ChatGPT

main_agent — Claude Code prompt: what it does and how to install

This Claude Code prompt delegates visual observation while leaving physical actions and task completion to the main agent. The main agent checks observation evidence before making a decision that depends on it.

MIT ★ 488 updated 2026-11-15

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
.claude/commands/main_agent.md
Open on GitHub

What main_agent does

The prompt assigns an observer one focused visual question, along with task context and a condition for stopping. The main agent passes along the observer instructions, then reads results through get_latest_insight or read_insight_updates and tracks which insights it has consumed.

Before acting on an observation, the main agent checks its snapshot version, freshness, findings and uncertainty. A snapshot from an unfinished action does not establish that the action has finished, and evidence captured before an action cannot confirm its effect. If evidence is missing or calibrated geometry is needed, the prompt calls for view_env_state or capture_scene. The observer does not control physical actions or decide when the task is complete.

What it helps with

  • delegate a focused visual question to an observer
  • check observation freshness before a dependent decision
  • avoid treating a snapshot from an unfinished action as proof of completion
  • keep physical control and the final task decision with the main agent

Example prompts for the agent

  • Check the scene for a visible change in the object’s position and report whether it is clear.
  • Compare the latest observation with the earlier one and describe the visible changes.
  • Review the visual evidence before the next action; request a new snapshot if the evidence is unclear.

How to install main_agent

Manually: put the files in place

.claude/commands/main_agent.md

Questions about main_agent

How do I install main_agent in Claude Code?

Save the prompt as .claude/commands/main_agent.md.

What does main_agent do?

It sets out how to delegate a visual question and inspect the resulting evidence. The main agent remains responsible for physical actions and task completion.

What can the observer handle?

It can handle a specific visual question. Physical control, metric geometry derivation and the completion decision stay with the main agent.

Source text

main_agent.md · MIT
# Observation Sidecar

The observation profile shares versioned snapshots and visual insights. The
main agent retains responsibility for geometry, physical actions, recovery and
the final task decision.

When delegation is available and useful, assign one concrete visual question.
Supply the task context, a distinct observer_id, an optional baseline
observation_id, a visually decidable stopping condition and the injected
observation_subagent_instructions block. Forward that block with the assignment;
it describes the observer's role, not yours. Delegate observation only, without
physical control, metric geometry derivation or finish.

Read get_latest_insight or read_insight_updates, tracking the last consumed
insight_seq per observer and advancing the returned cursor. Lifecycle messages
do not replace an insight tied to observation_id and world_version. Continue
independent work, then read relevant evidence before a dependent decision.

Check the insight's source snapshot/state version, freshness, facts, changes,
risks and uncertainty against the action being assessed. Published snapshots
can come from completed primitives within an unfinished high-level action;
in_flight does not prove that action has completed. A fresh tag alone does not
prove the desired effect. Evidence from before an action cannot verify its effect.

Use view_env_state or capture_scene when evidence is missing, ambiguous or too
old, or calibrated geometry is needed. Avoid repeating a resolved visual check
unless the scene changed. End the assignment once its question is resolved;
the main agent retains the completion decision and calls finish.

© Open-RUA · MIT · Open-RUA/uni-robot-agent

Similar prompts

Prompts catalog →
rebase-latest-main dyoshikawa rebase-latest-main — a skill that walks the agent through fetching origin/main, rebasing the current branch and resolving conflicts. Handy when your branch has drifted behind main. CX Codex +1 ★ 1,636repo stars pull-main dyoshikawa pull-main is a Claude Code skill that switches to main and fetches its latest changes. It runs when asked to update or sync the branch with the remote. CX Codex +1 ★ 1,636repo stars vstack-main Hao AI Lab vstack-main — a subagent for Claude Code coordinates Verification Stack stages and checks their evidence independently. It accepts a stage only after reviewing its artifacts and results. CX Codex ★ 89repo stars cache-miss.sequence tt-a1i cache-miss.sequence drives the Archify skill to build a sequence diagram for a dashboard request that misses cache in Redis and falls through to PostgreSQL. CX Codex GP ChatGPT ★ 98,599repo stars web-runtime.architecture tt-a1i web-runtime.architecture — a prompt for generating a production web architecture JSON diagram using Archify and saving it directly to benchmark-candidate.json. CX Codex GP ChatGPT ★ 98,599repo stars product-analytics.dataflow tt-a1i This prompt guides Archify through a product analytics data-flow diagram, from web and mobile events to dashboards and a model, with a distinct path for consent-gated identity data. CX Codex GP ChatGPT ★ 98,599repo stars agent-tool-call.workflow tt-a1i agent-tool-call.workflow — a prompt for building a workflow diagram where an AI agent calls tools with human approval gates. CX Codex GP ChatGPT ★ 98,599repo stars agent-run.lifecycle tt-a1i agent-run.lifecycle instructs an agent to build a structured lifecycle diagram for an agent run using the Archify workflow. CX Codex GP ChatGPT ★ 98,599repo stars

From the same set41

The whole set →
hybrid The hybrid prompt for Claude Code reports RoboDojo task progress with evidence-backed subgoals. A reviewer checks action-model proposals before execution. GP ChatGPT ★ 488repo stars hybrid The hybrid Claude Code prompt controls an arm and dexterous hand assembly in DexJoCo. It supports pick_bucket with the published single-arm model recipe. GP ChatGPT ★ 488repo stars hybrid The hybrid prompt for Claude Code directs R1Pro through actor_act. It covers bounded local phases using the robot’s whole body and grippers. GP ChatGPT ★ 488repo stars return_background The return_background Claude Code prompt coordinates background jobs and waits. It defines queue boundaries, result checks, and recovery after failures. GP ChatGPT ★ 488repo stars hybrid The hybrid Claude Code prompt helps choose between actor_act and direct robot motion. It guides a bounded control phase and the checks that follow. GP ChatGPT ★ 488repo stars hybrid The hybrid prompt for Claude Code and ChatGPT guides a Go2 quadruped through a prepared indoor navigation episode. The agent chooses movement from RGB views and ends the run only by calling finish. GP ChatGPT ★ 488repo stars meta_skill The meta_skill Claude Code prompt guides agents through finding existing skills and recording lessons backed by a task. It comes into play when an agent needs to add a lesson or revise a skill. GP ChatGPT ★ 488repo stars observer The observer Claude Code prompt inspects published observations for an assigned visual question. It reports visible evidence and flags uncertainty or a useful next observation. GP ChatGPT ★ 488repo stars

Project sponsors

Companies that keep this analytics open