# AI use cases for Agentplay (Entelligence-AI/agentplay)

Repo: [Entelligence-AI/agentplay](https://github.com/Entelligence-AI/agentplay) | Use cases: 4 | Page: https://dedupedai.com/use-cases/entelligence-ai-agentplay/

A local, open source tool that replays coding agent sessions as a map and a timeline, showing what the agent read, edited, and retried.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Check what an agent read before it edited a file: https://dedupedai.com/use-cases/entelligence-ai-agentplay/check-what-agent-read-before-editing/
- Find the commands an agent kept retrying and failing: https://dedupedai.com/use-cases/entelligence-ai-agentplay/find-repeated-failing-commands/
- Share a session map without prompt text or file contents: https://dedupedai.com/use-cases/entelligence-ai-agentplay/share-session-map-without-prompt-text/
- Review what each Claude Code subagent brought back: https://dedupedai.com/use-cases/entelligence-ai-agentplay/review-subagent-work-in-claude-code/

## How to check what an agent read before it edited a file with Agentplay

Repo: [Entelligence-AI/agentplay](https://github.com/Entelligence-AI/agentplay) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Replay a session to see which files the agent opened before changing code, and which it never opened.

**Who it is for:** A developer reviewing an agent's pull-ready changes who wants to confirm it looked at the right code first.

**The problem:** A diff shows the final edit but not the investigation that led to it, so a reviewer cannot tell whether the agent read the relevant files.

**The outcome:** A clear list of files read, edited, and skipped in the session, with the order they were touched.

### Steps

1. Run `uvx agentplay` in a terminal to build maps and open the local viewer.
2. Open the session you want to review from the session list in the viewer.
3. Use the map to find the edited file and check which files were seen or read before it.
4. Press Space to play the timeline and watch the order of reads and edits.

### Prompt to paste into your agent

```text
In the agentplay repo at https://github.com/Entelligence-AI/agentplay, run `uvx agentplay` and open the session that edited <target file path>. Report which files the agent read before the first edit to that file and which related files it never opened. Verify by confirming each listed read appears earlier than the first edit on the timeline.
```

### Caveat

The map shows files and actions, not the agent's reasoning or the text of its prompts, so you still need the diff to judge the change itself.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/entelligence-ai-agentplay/check-what-agent-read-before-editing/

## How to find the commands an agent kept retrying and failing with Agentplay

Repo: [Entelligence-AI/agentplay](https://github.com/Entelligence-AI/agentplay) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Use the Errors tab to see which failing command ran most often in a row and jump to it in the replay.

**Who it is for:** A team lead who sees an agent stall on the same error and wants to know where it got stuck.

**The problem:** Long sessions hide repeated failures inside many tool calls, so it is hard to see which error wasted the most time.

**The outcome:** A ranked list of error classes with the longest run of the same failing command, and a replay position for each.

### Steps

1. Run `uvx agentplay` to build maps and open the viewer.
2. Open the session and select the Errors tab.
3. Read the list, which is grouped by error class and ranked by the longest run of the same failing command.
4. Click the top row to seek the replay to that error and see the surrounding steps.

### Prompt to paste into your agent

```text
Using the agentplay repo at https://github.com/Entelligence-AI/agentplay, run `uvx agentplay` and open the session <session name or id>. Open the Errors tab, list the top three error classes with their longest repeated failing command, and state the step where each run began. Verify that each count matches the error rows shown in the Errors tab.
```

### Caveat

Command arguments are not stored, so the Errors tab identifies commands by verb such as pytest or git log, not by their full arguments.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/entelligence-ai-agentplay/find-repeated-failing-commands/

## How to share a session map without prompt text or file contents with Agentplay

Repo: [Entelligence-AI/agentplay](https://github.com/Entelligence-AI/agentplay) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Build a path-redacted map and export one session as a file that carries no prompts or file contents.

**Who it is for:** An engineer who wants to post or send an example of an agent session to a colleague or a public issue without exposing code or prompts.

**The problem:** Raw transcripts contain prompts, tool outputs, and file contents that should not leave a machine.

**The outcome:** A single map file with repo names and redacted paths that a reader can inspect without seeing prompt text, tool outputs, or file contents.

### Steps

1. Run `agentplay build --path-mode dirs_only` to rebuild maps with directory-only paths. (inferred, check the README)
2. Run `uvx agentplay` to open the viewer and confirm the session appears in the list.
3. Run `agentplay export <session> <file>.json` to write one session's map to a file. (inferred, check the README)
4. Open the exported file and confirm it contains no prompt text or file contents before sharing.

### Prompt to paste into your agent

```text
In the agentplay repo at https://github.com/Entelligence-AI/agentplay, rebuild maps with `--path-mode dirs_only`, then export the session <session id> to <output file>.json. Report the repo names and the number of paths the export tool says it contains. Verify that the exported file has no prompt, assistant, or tool output text.
```

### Caveat

Repo names and remotes stay in every path mode, so remove the repos remote field before sharing if the remote is sensitive.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/entelligence-ai-agentplay/share-session-map-without-prompt-text/

## How to review what each Claude Code subagent brought back with Agentplay

Repo: [Entelligence-AI/agentplay](https://github.com/Entelligence-AI/agentplay) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Use subagent lanes and the agent lens to see how each subagent in a Claude Code session contributed.

**Who it is for:** A developer who runs Claude Code with subagents and wants to know which ones returned useful results.

**The problem:** Subagent work is easy to lose in a long session, so it is unclear which subagents contributed and which came back empty.

**The outcome:** A per-subagent view of the session showing each subagent's work in its own lane and its position in the timeline.

### Steps

1. Run `uvx agentplay` to build maps and open the viewer.
2. Open a Claude Code session that used subagents.
3. Select a single subagent with the agent lens so the explorer re-scopes to it.
4. Play the timeline to see what that subagent read, edited, and produced.

### Prompt to paste into your agent

```text
Using the agentplay repo at https://github.com/Entelligence-AI/agentplay, run `uvx agentplay`, open the Claude Code session <session name or id>, and use the agent lens to scope to each subagent. Report which subagents produced edits or outputs and which came back with nothing. Verify by checking each subagent lane against its entries on the timeline.
```

### Caveat

Subagent prompts and descriptions are not recorded, so you can see what each subagent did but not the instructions it was given.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/entelligence-ai-agentplay/review-subagent-work-in-claude-code/