Skip to main content
← Back to Prompt Modifiers

Prompt Modifiers

Output Format Variants (Modifier)

Best for
When you need audit findings in a specific format for your workflow
Use when
Creating tickets, PR comments, Slack summaries, or executive reports

Append this to any other prompt and specify which output format you need.

Available Formats

PR Comments — One finding per comment, ready to paste into a GitHub/GitLab review (or deliver directly: gh pr review --comment, gh issue create when asked to file them):

**[SEVERITY]** Issue title

`file:line` — Description of the issue.

**Suggested fix:**
\`\`\`language
// code fix here
\`\`\`

Tickets (Jira/Linear/GitHub Issues) — Each finding as a standalone ticket:

**Title:** [SEVERITY] Brief issue description
**Labels:** security | performance | ux | bug
**Priority:** P1/P2/P3/P4 (maps to the library severity scale: P1=Critical, P2=High, P3=Medium, P4=Low)

**Description:**
What's wrong and why it matters.

**Acceptance Criteria:**
- [ ] Specific condition that must be true when fixed
- [ ] Test case that verifies the fix

**Technical Notes:**
File locations, code context, and implementation guidance.

Slack Summary — Concise message for a channel update:

*[Audit Name] Results*
Health: [emoji] [Good/Needs Attention/Urgent]
Found: X critical, Y high, Z medium

Top 3 actions:
1. [Critical] One-line description — `file`
2. [High] One-line description — `file`
3. [High] One-line description — `file`

Full report: [link or thread]

Executive Summary — For non-technical stakeholders:

## [Audit Name] — Summary

**Overall Assessment:** [1-2 sentences on health and risk level]

**Key Risks:** [2-3 bullet points in business language, no code]
**Key Strengths:** [1-2 bullet points on what's done well]
**Recommended Actions:** [2-3 prioritized next steps with effort estimates]
**Timeline:** [Suggested order and urgency]

Changelog / ADR — For documenting decisions made based on findings:

## Decision: [What was decided]
**Date:** [date]
**Context:** [Audit finding that prompted this]
**Decision:** [What we're doing about it]
**Consequences:** [Trade-offs accepted]
**Status:** Proposed | Accepted | Implemented

Usage

Append to any prompt: "Format output as: [PR Comments | Tickets | Slack Summary | Executive Summary | Changelog]"

You can request multiple formats: "Format output as: Slack Summary for the team channel, then Tickets for the top 5 findings."

JSON (machine-readable) — For piping into scripts or dashboards, an array of finding objects:

[{ "file": "path/to/file.ts", "line": 42, "severity": "Critical|High|Medium|Low", "confidence": "Confirmed|Likely|Speculative", "summary": "one-sentence defect", "fix": "one-sentence remedy" }]

Emit valid JSON only (no prose around it) when this format is requested.

Need help applying this to a real product?

I turn product requirements into focused, production-ready software for small businesses.