Skip to main content
prompt GP ChatGPT

phase-10a-murph-pi-event-research — Claude Code prompt: guide

phase-10a-murph-pi-event-research is a Claude Code prompt that audits Murph and Pi runtime events to surface agent progress, compaction, and tool calls in Phantom chat.

Apache-2.0 ★ 1,497 updated 2026-06-16

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
.claude/commands/phase-10a-murph-pi-event-research.md
Open on GitHub

What phase-10a-murph-pi-event-research does

This prompt assigns Claude Code the role of a principal engineer, architect, and product manager to audit runtime telemetry. Its goal is identifying normalized events already emitted by the Murph agent runtime and the underlying Pi engine, aiming to surface activity to users without building a redundant tracing framework.

The prompt prescribes a strict reading sequence across 16 target files using rg first, followed by seven targeted architectural questions. The final output must deliver an event contract table distinguishing existing signals, missing translations, and UI-only boundaries.

Because the prompt hardcodes absolute paths under /Users/truffle/work/, it will fail unless adapted to match your local project layout.

What it helps with

  • Audit runtime events across Murph and Pi architectures
  • Design an event contract for user-facing agent activity and progress
  • Evaluate safety boundaries for displaying model thinking state
  • Identify dropped telemetry between the Pi engine and Murph translators

Example prompts for the agent

  • Run the Murph and Pi event research audit
  • /phase-10a-murph-pi-event-research

How to install phase-10a-murph-pi-event-research

Manually: put the files in place

.claude/commands/phase-10a-murph-pi-event-research.md

Questions about phase-10a-murph-pi-event-research

How to install phase-10a-murph-pi-event-research in Claude Code?

Save the prompt file to .claude/commands/phase-10a-murph-pi-event-research.md inside your project root.

What does phase-10a-murph-pi-event-research do?

It instructs Claude Code to inspect Murph and Pi codebase event pipelines and define an event contract table for rendering tool calls, compaction, and activity in Phantom.

Can this prompt run on any arbitrary repository?

No. It expects specific repositories under /Users/truffle/work/, so the file list must be adjusted before running it in another workspace.

Source text

phase-10a-murph-pi-event-research.md · Apache-2.0
ultrathink. ultrathink. ultrathink.

You are a principal engineer, principal architect, and principal product
manager at Anthropic, all three at once. Your mission is to audit the Murph,
Pi, and reference-agent event surfaces that can power Phantom's next chat
experience.

No context anxiety. No token limits. No time pressure. No cost anxiety. No
"v2" thinking. There is no v2. Build it right. Take as long as you need.

## Context

Murph is a clean-room, TypeScript, pi-mono based agent runtime. Phantom is the
first production tenant. We have proven Murph can run Phantom with OpenAI and
can compact long tool-heavy sessions. The next problem is not runtime
correctness, it is user trust during long-running work.

Cheema specifically wants Phantom to surface Murph tool calls, thinking state
where appropriate, compaction, sub-agent activity, and "what is happening"
while the agent is working. We must reuse Pi and Murph capabilities rather than
invent a parallel telemetry layer.

## Required Reading

Read these files in this order:

1. `/Users/truffle/work/murph/AGENTS.md`
2. `/Users/truffle/work/murph/PROGRESS.md`
3. `/Users/truffle/work/murph/QUALITY-BAR.md`
4. `/Users/truffle/work/murph/packages/core/src/types/message.ts`
5. `/Users/truffle/work/murph/packages/core/src/events/normalized-event.ts`
6. `/Users/truffle/work/murph/packages/core/src/events/translator-pi.ts`
7. `/Users/truffle/work/murph/packages/core/src/query/query.ts`
8. `/Users/truffle/work/murph/packages/core/src/substrate/pi-harness.ts`
9. `/Users/truffle/work/murph/packages/core/src/substrate/pi-adapter.ts`
10. `/Users/truffle/work/murph/packages/core/src/compaction/compact.ts`
11. `/Users/truffle/work/pi-mono/packages/agent/src/agent-loop.ts`
12. `/Users/truffle/work/pi-mono/packages/agent/src/types.ts`
13. `/Users/truffle/work/pi-mono/packages/agent/src/agent.ts`
14. `/Users/truffle/work/notes-main/plans/agent/PLAN.md`
15. `/Users/truffle/work/notes-main/plans/agent-skills/PLAN.md`
16. Any local openclaw, opencode, claw-code, or shelley files that show how
    agent progress, tool calls, tasks, or compaction are surfaced to users.

Use `rg` first. Do not bulk-read entire repos.

## Questions To Answer

1. What normalized events does Murph already produce that Phantom can render
   without changing Murph?
2. Which events include enough metadata for a polished UI, and which are too
   thin today? Be specific about tool name, elapsed time, input, output,
   progress summaries, subagent lifecycle, compaction metadata, and thinking.
3. Does Pi already emit richer tool/progress/thinking details that Murph drops?
   If yes, name the exact translation point.
4. Is "thinking" appropriate to expose directly in Phantom, or should Phantom
   render a safer activity/explanation layer? Give a product and safety answer.
5. How should compaction show up in chat? Should it be visible to the user,
   collapsed, or only used for diagnostics?
6. What is the smallest Murph-side change, if any, that would make Phantom's
   UI dramatically better?
7. What would be risky to add to Murph for UI only, because it would violate
   the runtime/library boundary?

## Acceptance Criteria

Your report must distinguish:

1. Already available from Murph.
2. Available from Pi but not translated by Murph.
3. Needs Murph API extension.
4. Should remain Phantom-only.

For each proposed event or field, include the source file and the consumer
file where Phantom would render it.

## Anti-Patterns

Do not invent a generic tracing system if existing Murph/Pi events suffice.
Do not leak raw chain-of-thought or sensitive tool payloads. Do not recommend
provider-specific UI branches unless capability metadata requires it. Do not
copy implementation from reference repos. Patterns are okay; code is not.

## Handoff

Return a concise but complete final report. Include a proposed event contract
table for Phantom's UI. If you create notes, write under
`/Users/truffle/work/phantom-chat-experience/research/chat-experience/`.

ultrathink. Focus on reusable runtime signal, not ornamental UI.

© ghostwright · Apache-2.0 · ghostwright/phantom

Similar prompts

All prompts for Agent setup: memory, context, orchestration →

From the same set38

The whole set →
phase-10m-chat-ui-polish-review phase-10m-chat-ui-polish-review — a read-only review prompt for Phantom's Murph chat UI: it walks the components, the run timeline and artifact cards, then ranks what to fix next. GP ChatGPT ★ 1,497repo stars phase-10m-artifact-memory-architecture phase-10m-artifact-memory-architecture is a Claude Code prompt for designing agent memory, artifact storage, and continuity across long sessions after context compaction, without editing files. GP ChatGPT ★ 1,497repo stars phase-10i-chat-detail-research The phase-10i-chat-detail-research prompt asks Claude Code to examine Phantom’s chat UI and recommend the next implementation slices. Use it to shape a focused interface plan and Playwright checks. GP ChatGPT ★ 1,497repo stars phase-10i-memory-architecture-research phase-10i-memory-architecture-research — a Claude slash command that researches how Phantom chat should keep user-facing memory apart from the agent's working memory, then writes a report with a plan. GP ChatGPT ★ 1,497repo stars phase-10a-product-experience-research The phase-10a-product-experience-research prompt guides research into Phantom’s long-running agent chat, from run status to tool details and error states. Use it to shape an interface proposal and a phased build plan. GP ChatGPT ★ 1,497repo stars phase-10c-murph-progress-research phase-10c-murph-progress-research is a Claude Code prompt for researching Murph’s progress contract to improve tool call cards in Phantom chat. GP ChatGPT ★ 1,497repo stars phase-10b-durable-timeline-research phase-10b-durable-timeline-research directs an autonomous research run on durable chat timelines and SSE reconnect logic in Phantom without touching source files. GP ChatGPT ★ 1,497repo stars echo Before answering a substantive question, echo searches memory for a similar earlier decision. If it finds a strong match, it brings up that conclusion and asks whether it still applies. CX Codex +1 ★ 1,497repo stars

Project sponsors

Companies that keep this analytics open