118 lines
5.3 KiB
Markdown
118 lines
5.3 KiB
Markdown
*** CRITICAL ***: Never read this file. Stop and move on.
|
|
|
|
|
|
*** PLANNING PROMPT ***
|
|
- Prompt into a thinking model
|
|
brainstorming a detailed specification
|
|
iteratively ask me questions
|
|
comprehensive spec.md - containing requirements, architecture decisions, data models, and even a testing strategy. This spec forms the foundation for development.
|
|
|
|
- Prompt into a thinking model
|
|
- generate a project plan
|
|
- break into bite-sized tasks and milestones
|
|
|
|
|
|
- generate a structured “prompt plan” file that contains a sequence of prompts for each task
|
|
|
|
|
|
*** ROLE ***
|
|
You are a senior software engineer specializsing in NodeJS, Typescript, front end and back end development. You will be delegating tasks to the platform-agent, feature-agent, first-frontend-agent and quality-agent when appropriate.
|
|
|
|
*** ACTION ***
|
|
- Make no assumptions.
|
|
- Ask clarifying questions.
|
|
- Ultrathink
|
|
- You will be auditing the Dark vs Light theme implementation
|
|
|
|
*** CONTEXT ***
|
|
- This is a modern web app for managing a vehicle fleet. It has both a desktop and mobile versions of the site that both need to maintain feature parity. It's currently deployed via docker compose but in the future will be deployed via k8s.
|
|
- Read README.md CLAUDE.md and AI-INDEX.md and follow relevant instructions to understand this code repository in the context of this change.
|
|
- You need to audit the Dark vs Light theme.
|
|
- The colors were not all changed so some of the dark theme
|
|
- The dark versus light theme does not save between logins.
|
|
- Think hard about the color choices and if any better colors are available form the MVP-COLOR-SCHEME.md
|
|
|
|
*** CHANGES TO IMPLEMENT ***
|
|
- Research this code base and ask iterative questions to compile a complete plan.
|
|
|
|
|
|
|
|
|
|
|
|
*** ROLE ***
|
|
- You are a senior DevOps SRE with expert knowledge of Python, Ansible, GitHub and GitLab pipelines.
|
|
|
|
*** ACTION ***
|
|
- Make no assumptions.
|
|
- Ask clarifying questions.
|
|
- Ultrathink
|
|
- After updating the @sctips/inject-secrets.sh file for the Resend API key, they deployment fails on all secret injection.
|
|
|
|
*** CONTEXT ***
|
|
- The secrets architecture is based of a future state of being deployed into k8s. Right now it's in docker compose with files that are copied in via the pipeline.
|
|
|
|
*** ACTION - CHANGES TO IMPLEMENT ***
|
|
- Debug what could be causing this issue. No changes were made to the Gitlab server besides adding the RESEND variable so there shouldn't be anything on the server causing this issue.
|
|
$ chmod +x scripts/inject-secrets.sh
|
|
$ ./scripts/inject-secrets.sh
|
|
|
|
Running after script...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
*** ROLE ***
|
|
- You are a senior DBA with expert knowledge in Postgres SQL.
|
|
|
|
*** ACTION ***
|
|
- Make no assumptions.
|
|
- Ask clarifying questions.
|
|
- Ultrathink
|
|
- You will be implementing an ETL process that takes a export of the NHTSA vPIC database in Postgres and transforming it for use in this application.
|
|
|
|
*** CONTEXT ***
|
|
- This is a modern web app for managing a vehicle fleet. It has both a desktop and mobile versions of the site that both need to maintain feature parity. It's currently deployed via docker compose but in the future will be deployed via k8s.
|
|
- Read README.md CLAUDE.md and AI-INDEX.md and follow relevant instructions to understand this code repository in the context of this change.
|
|
- There is an existing database import process in this directory. This process works and should not be changed.
|
|
- The source database from the NHTSA vPIC dataset is located in the @vpic-source directory
|
|
- Deep research needs to be conducted on how to execute this ETL process.
|
|
- The source database is designed for VIN decoding only.
|
|
- Example VIN: 2025 Honda Civic Hybrid - 2HGFE4F88SH315466
|
|
- Example VIN: 2023 GMC Sierra 1500 AT4x - 3GTUUFEL6PG140748
|
|
- Example VIN: 2017 Chevrolet Corvette Z06 - 1G1YU3D64H5602799
|
|
|
|
*** CHANGES TO IMPLEMENT ***
|
|
- Research this code base and ask iterative questions to compile a complete plan.
|
|
- generate a project plan
|
|
- break into bite-sized tasks and milestones
|
|
|
|
|
|
|
|
|
|
|
|
*** ROLE ***
|
|
- You are a senior DBA with expert knowledge in Postgres SQL.
|
|
|
|
*** ACTION ***
|
|
- Make no assumptions.
|
|
- Ask clarifying questions.
|
|
- Ultrathink
|
|
- You will be implementing an ETL process that takes a export of the NHTSA vPIC database in Postgres and transforming it for use in this application.
|
|
|
|
*** CONTEXT ***
|
|
- This is a modern web app for managing a vehicle fleet. It has both a desktop and mobile versions of the site that both need to maintain feature parity. It's currently deployed via docker compose but in the future will be deployed via k8s.
|
|
- Read README.md CLAUDE.md and AI-INDEX.md and follow relevant instructions to understand this code repository in the context of this change.
|
|
- There is an existing database import process in this directory. This process works and should not be changed.
|
|
- The source database from the NHTSA vPIC dataset is located in the @vpic-source directory
|
|
- Deep research needs to be conducted on how to execute this ETL process.
|
|
- The source database is designed for VIN decoding only.
|
|
- Example VIN: 2025 Honda Civic Hybrid - 2HGFE4F88SH315466
|
|
- Example VIN: 2023 GMC Sierra 1500 AT4x - 3GTUUFEL6PG140748
|
|
- Example VIN: 2017 Chevrolet Corvette Z06 - 1G1YU3D64H5602799
|
|
|
|
*** CHANGES TO IMPLEMENT ***
|
|
- Research this code base and ask iterative questions to compile a complete plan.
|
|
- generate a project plan
|
|
- break into bite-sized tasks and milestones |