Prompt Library
Prompts for coding, docs, automation and Salesforce work, in Markdown and JSON.

Who it's for
- You use LLMs for engineering work every day and keep rewriting the same prompts for reviews, tests and docs.
- You want prompts that say exactly what context they need and what shape the answer must have, including JSON answers with a schema.
- You build automations and want prompts you can load from a file instead of pasting them into nodes by hand.
Not for you if you want short, general-purpose prompts: the Prompt Cheat Sheets are the one-page references.
What's inside
88 prompts in nine categories: code review and refactoring, debugging and incidents, tests and test data, documentation, design and ADRs, extraction and transformation, agents and n8n automation design, communication (commits, PRs, release notes, client emails) and Salesforce (Apex, LWC, SOQL, Flow and deployment). Every prompt states the context it needs, its variables and its output format; the prompts that answer in JSON come with a JSON Schema for the answer.
The whole collection is one JSON file, with a readable Markdown page per prompt generated from it. Three scripts fill a prompt's variables from your own files, regenerate the pages after you edit the JSON, and export to CSV, one Markdown file or snippet-manager JSON. The scripts and tests use only Node's standard library.
Each prompt will also ship with a saved run (the exact input, the model, the date and the raw answer). Those runs are being recorded now and are not in the current build.
- Prompts (Markdown)
- 88
- prompts/architecture/adr-from-discussion.md
- prompts/architecture/api-design-review.md
- prompts/architecture/build-vs-buy.md
- prompts/architecture/capacity-estimate.md
- prompts/architecture/compare-options-json.md
- prompts/architecture/data-model-from-requirements.md
- prompts/architecture/design-doc-review.md
- prompts/architecture/failure-modes-json.md
- prompts/architecture/migration-plan.md
- prompts/automation/agent-guardrails.md
- prompts/automation/answer-from-documents.md
- prompts/automation/eval-cases-for-prompt.md
- prompts/automation/map-webhook-payload.md
- prompts/automation/n8n-code-node.md
- prompts/automation/process-to-automation-spec.md
- prompts/automation/review-agent-transcript.md
- prompts/automation/route-request.md
- prompts/automation/write-agent-system-prompt.md
- prompts/automation/write-tool-definition.md
- prompts/code-review/refactor-extract-functions.md
- prompts/code-review/refactor-plan-safe-steps.md
- prompts/code-review/review-dependency-upgrade.md
- prompts/code-review/review-diff-json-findings.md
- prompts/code-review/review-error-handling.md
- prompts/code-review/review-performance.md
- prompts/code-review/review-pull-request-diff.md
- prompts/code-review/review-readability.md
- prompts/code-review/review-sql-query.md
- prompts/code-review/security-review-code.md
- prompts/communication/changelog-entries.md
- prompts/communication/client-email-delay.md
- prompts/communication/client-email-scope-change.md
- prompts/communication/commit-message.md
- prompts/communication/explain-technical-decision.md
- prompts/communication/incident-customer-update.md
- prompts/communication/pull-request-description.md
- prompts/communication/rewrite-review-comments.md
- prompts/communication/weekly-status-update.md
- prompts/data/compare-json-payloads.md
- prompts/data/extract-contacts.md
- prompts/data/extract-invoice-fields.md
- prompts/data/extract-requirements.md
- prompts/data/extract-system-dependencies.md
- prompts/data/normalize-records.md
- prompts/data/parse-log-lines.md
- prompts/data/question-to-sql.md
- prompts/data/redact-personal-data.md
- prompts/data/triage-support-ticket.md
- prompts/debugging/bug-report-to-hypotheses.md
- prompts/debugging/diagnose-flaky-test.md
- prompts/debugging/diagnose-stack-trace.md
- prompts/debugging/hotfix-risk-check.md
- prompts/debugging/incident-timeline.md
- prompts/debugging/investigate-memory-leak.md
- prompts/debugging/investigate-performance-regression.md
- prompts/debugging/minimal-reproduction.md
- prompts/debugging/postmortem-draft.md
- prompts/debugging/triage-log-excerpt.md
- prompts/docs/api-endpoint-reference.md
- prompts/docs/docs-gap-audit.md
- prompts/docs/docstrings-for-module.md
- prompts/docs/edit-for-clarity.md
- prompts/docs/explain-module-onboarding.md
- prompts/docs/faq-from-support-threads.md
- prompts/docs/readme-from-facts.md
- prompts/docs/runbook-from-notes.md
- prompts/docs/tutorial-from-verified-steps.md
- prompts/docs/upgrade-guide.md
- prompts/salesforce/apex-code-review.md
- prompts/salesforce/apex-test-class.md
- prompts/salesforce/apex-trigger-handler.md
- prompts/salesforce/deploy-error-triage.md
- prompts/salesforce/flow-design.md
- prompts/salesforce/flow-or-apex.md
- prompts/salesforce/integration-pattern.md
- prompts/salesforce/lwc-component.md
- prompts/salesforce/permission-set-design.md
- prompts/salesforce/soql-review.md
- prompts/testing/api-contract-tests.md
- prompts/testing/boundary-values-from-rules.md
- prompts/testing/e2e-scenarios-gherkin.md
- prompts/testing/find-test-gaps-mutation.md
- prompts/testing/integration-test-plan.md
- prompts/testing/property-based-tests.md
- prompts/testing/review-test-file.md
- prompts/testing/test-cases-from-requirement.md
- prompts/testing/test-fixtures-from-schema.md
- prompts/testing/unit-tests-for-function.md
- Prompt collection (JSON)
- 1
- prompts.json
- Scripts (fill, render, export)
- 3
- scripts/export.mjs
- scripts/fill.mjs
- scripts/render-prompts.mjs
- Guides
- 4
- docs/how-to-use.md
- docs/how-we-tested.md
- docs/in-your-tools.md
- docs/variables.md
Files
Files
Files
Files
In the zip · 113 files · 291 KB
- CHANGELOG.md
- LICENSE.md
- README.md
- docs/4 files
- package-lock.json
- package.json
- prompts.json
- prompts/89 files
- schema/1 files
- scripts/5 files
- tests/8 files
Setup
Requirements
- Any LLM chat tool or API (Claude, ChatGPT, Gemini or similar)
- Node.js 22 or newer only if you use the fill-in, render and export scripts or the tests (no dependencies)
- Open
prompts/index.md, pick a prompt (for example "Review a pull request diff") and copy its Prompt block. - Replace each
{{variable}}with your own text and paste it into Claude, ChatGPT or another model. - Or use the fill-in script:
node scripts/fill.mjs --list, thennode scripts/fill.mjs review-pull-request-diff --var diff=@change.diff > filled.txt. - After you edit
prompts.json, runnpm testto check the collection.
Time to first result: about 3 minutes.
Tested with
The clean-install test record for this exact zip is still to come.
Licence
You may use this product in unlimited personal and commercial projects, including client work. You may not resell, redistribute, sublicense or share the product itself, in whole or in part, including as a template, starter kit or course material. One purchase covers one person.
The full licence is in LICENSE.md inside the zip. Site terms: /terms.
Questions
When can I buy it?
When every prompt has a recorded test run with a real model and the screenshots are retaken. I haven't set a date; leave your email above and I'll tell you when it's on sale.
Which models do the prompts work with?
Any capable chat model: Claude, ChatGPT, Gemini or similar. The saved runs will name the exact model and date each prompt was tested with.
Do I need Node.js?
Only for the scripts and tests (Node.js 22 or newer, nothing to install). Without it you copy prompts from the Markdown pages.
How do updates work?
New versions go to your Gumroad library; download them from there. Changes are listed in the CHANGELOG inside the zip.
Can I use it for client work?
Yes, in unlimited personal and commercial projects, including client work. You can't resell or share the prompts as a prompt pack.