> ## Documentation Index
> Fetch the complete documentation index at: https://doc.entelligence.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Ask Ellie prompt library

> Copy scoped prompts for developers, engineering leaders, platform teams, incidents, code review, and AI-agent operations.

Replace the uppercase placeholders. Add “do not create or update anything” when you want a read-only answer.

## Developers

```text theme={null}
Explain how SERVICE handles REQUEST. Cite the repository paths and recent pull requests that changed the flow.
```

```text theme={null}
For PR_URL, summarize the behavior change, identify dependent codebases, and list the three highest-risk assumptions. Do not post or update anything.
```

```text theme={null}
Show recurring review findings for OWNER/REPO in the last 30 days. Group them by theme and link representative pull requests.
```

## Engineering leaders

```text theme={null}
For TEAM during DATE_RANGE, compare time to first review, cycle time, comment acceptance, and merged pull requests with the previous period. Explain likely drivers and include source links.
```

```text theme={null}
Which PROJECTS are over or near their AI-agent budget this month? Show spend, run rate, outcome mix, and any untracked spend.
```

```text theme={null}
Summarize AI-agent adoption and context-health changes for TEAM this month. Separate data gaps from actual declines.
```

## Platform and developer experience

```text theme={null}
Find repositories with repeated churn hotspots, long review cycles, and frequent agent retries. Show the evidence for each repository.
```

```text theme={null}
Which members or teams are unsynced across Cursor, Claude Code, Codex, Linear, or Slack? Show the missing mapping, not private session content.
```

```text theme={null}
Create a draft weekly report instruction for code-review health, AI spend, and incident patterns. Do not schedule it yet.
```

## Reliability and incidents

```text theme={null}
For SERVICE, list active recurring incident patterns by severity and confidence. Include detection count, last activity, related pull requests, and fix-monitoring state.
```

```text theme={null}
Explain why PATTERN_ID is recurring. Cite representative detections and the repository areas involved. Do not trigger a fix.
```

```text theme={null}
Compare the latest observability audit with the previous audit for OWNER/REPO. Show score changes, open P0/P1 issues, and affected files.
```

## Follow-up patterns

* “What source and time range did you use?”
* “Show the underlying pull requests.”
* “Separate missing data from zero activity.”
* “Group by repository, then team.”
* “Draft the action, but do not execute it.”
* “What permission is required for that action?”
* “What will change if I confirm?”

## Next step

Read [permissions and actions](/ask-ellie/permissions) before enabling write-capable workflows.