ChatGPT + veriova

Give your Custom GPT shared context, rules, and optional reviews via veriova. No coding required-everything is done inside ChatGPT's GPT Builder.

What you'll get

After following this guide, your Custom GPT will be able to:

  • Load shared context at the start of work so your GPT is grounded in real project decisions
  • Recall past conventions and decisions when you ask about something discussed before
  • Inspect rules and run optional reviews before AI-built work is treated as ready for engineering
  • Share context across your team if multiple people use the same GPT

How it works (simple version)

veriova sits behind your GPT as the shared context and rules layer. When you tell ChatGPT something important, it can save that context to veriova. When you ask a question or review work, it can load context, inspect rules, and reuse what the team already knows.

ðŸ-¬

You chat with GPT

Tell it decisions, rules, or facts

🧠

GPT saves to veriova

Automatically stores the important bits

-Ϭ

GPT remembers forever

Recalls context in future conversations

Step 1: Sign up for veriova

Go to veriova.com/signup and enter your email. You'll receive a magic link-click it to verify.

Use OAuth auth in Actions.

Step 1.1: Create one OAuth client (one-time)

Create a client once and reuse it across GPT ID changes:

register-oauth-client.sh
curl -sS https://api.veriova.com/oauth/register \
  -H "Content-Type: application/json" \
  -d '{
    "client_name":"Veriova GPT",
    "redirect_uris":[
      "https://chat.openai.com/aip/g-any/oauth/callback",
      "https://chatgpt.com/aip/g-any/oauth/callback"
    ]
  }'

Save the returned client_id.

Step 2: Create your Custom GPT

In ChatGPT, go to Explore GPTs -†- Create. Switch to the Configure tab.

Name your GPT something descriptive, like “My Project Assistant” or “Team Memory Bot”.

Step 3: Add the Instructions

Paste the following into the Instructions box. This tells your GPT how to use veriova:

GPT Instructions
You are a project assistant with shared context powered by veriova.

## CONTEXT PROTOCOL

### Auto-Store Context (use the "pushMemory" action)
Automatically store memories when you:
- Make or discuss an important DECISION (type: "decision")
- Establish a CONVENTION or coding standard (type: "convention")
- Discover a reusable PATTERN (type: "pattern")
- Learn a LESSON from a mistake or debugging (type: "lesson")
- Write a useful code SNIPPET (type: "snippet")
- Reference external docs or URLs (type: "reference")

For each memory, choose a clear title and detailed content.
Set importance 1-10 (default 5, use 8+ for critical decisions).
Add relevant tags for categorization.

### Auto-Recall Context (use the "searchMemory" action)
Search memory when:
- The user starts a new topic or asks about a past decision
- You need to check if a convention already exists
- The user asks "what did we decide about X?"
- You're about to make a decision that might contradict a past one

### Context Packs (use the "generateContextPack" action)
Generate a full context pack when:
- Starting a new conversation or session
- The user says "catch me up" or "what do you know about X?"
- Working on a complex topic that may have multiple related memories

### Session Context (use the "getSessionContext" action)
- At the start of conversations where project context matters
- Before making implementation recommendations
- Before reviewing whether AI-built work is ready for engineering

### Rules + Optional Review
- Use "listStandards" before advising on production readiness
- Use "runReadinessReview" when the user asks whether a feature, prototype, or PR is ready to ship

## BEHAVIOR
- When recalling memories, cite them naturally: "Based on our previous decision..."
- Don't store trivial chat messages-only meaningful decisions, patterns, and learnings
- If you find conflicting memories, flag them to the user
- Proactively store important decisions without being asked
- At the start of each conversation, load context for the topic being discussed
- When discussing shipping or readiness, use saved rules and optional review instead of guessing

Step 4: Add the Actions

In the Actions section, click Create new action.

Authentication settings:

  • Type: OAuth
  • Client ID: cvo_...
  • Client Secret: not_used (or blank)
  • Authorization URL: https://api.veriova.com/oauth/authorize
  • Token URL: https://api.veriova.com/oauth/token
  • Scope: mcp
  • Token Exchange Method: POST request

Schema:

Paste the OpenAPI schema URL into the import field:

https://api.veriova.com/openapi-actions.yaml

Step 5: Test it

Save your GPT and start chatting. Try these prompts to see it in action:

"We decided to use PostgreSQL for the database. Remember this."
-†- GPT will call pushMemory automatically

"What database did we choose?"
-†- GPT will call searchMemory and cite the decision

"Catch me up on the project"
-†- GPT will call generateContextPack for a full briefing

"Is this payments flow ready for engineering continuation?"
-†- GPT will call listStandards and runReadinessReview

Step 6: Share with your team (optional)

Publish your GPT as “Anyone with the link” or “Only people in your workspace”.

For team use, create additional API keys with different roles from the API Keys dashboard:

  • Reader keys for team members who only need to search
  • Reviewer keys for those who should also store and verify memories
  • Admin keys for team leads who manage the project

Monitoring usage

Once your GPT is active, you can see how it's being used from the Dashboard. It shows:

  • API calls per day and which keys are being used
  • What endpoints are called (search, store, context pack)
  • Recent activity feed with timestamps
  • Memory breakdown by status and type

Tips for getting the most out of it

  • Be specific with decisions-“We chose React with TypeScript for the frontend because of team experience” is better than “Using React”.
  • Use tags-they help the AI categorize and find related memories later.
  • Review your memories-visit the Memory page to verify, pin important ones, and deprecate outdated ones.
  • Start each conversation with context --- say “catch me up” to get a briefing on everything the GPT knows.

Need help? Troubleshooting guide · API docs · Full docs