Lesson 4 of 6 — GitHub manifest connector

Automated Discovery Connectors · Source: Jira Story GOV-304 (5 pts) · Authority: Informational (the Jira story is authoritative).

Purpose

Discover model and agent manifests in GitHub repos.

Why it matters

Code repos hold undeclared models and agents — a common blind spot.

Where this lives

Runs in the Customer AI Governance platform (components: intelligence). The GitHub connector scans repos for model/agent manifests. Learn teaches this read-only.

Workflow

1. Given repositories
2. When the connector runs
3. Then model and agent manifests are discovered, closing the Git gap

Worked example

An AI agent is defined only in a repo manifest, never registered. How is it found?

The GitHub manifest connector discovers it, closing the Git discovery gap.

Acceptance criteria (authoritative — GOV-304)

Given repositories, when the connector runs, then model and agent manifests are discovered, closing the Git gap.

Common pitfalls

• Treating repos as out of scope and missing embedded AI.

Knowledge check

• Scenario: An AI agent is defined only in a repo manifest, never registered. How is it found?
• Look for: The GitHub manifest connector discovers it, closing the Git discovery gap.

Key concepts

• GitHub connector
• model/agent manifests
• Git gap

For the AI Tutor

• Summary: Discover model and agent manifests in GitHub repos.
• When to use: teaching or answering questions about "github manifest connector".
• When NOT to use: de-duplicating discovered systems (that is GOV-113).
• Key concepts: GitHub connector, model/agent manifests, Git gap.

Completion criteria

Learner passes the scenario knowledge check and can point to the system state that satisfies GOV-304's acceptance criterion.

Related

• Module: LEARN-MOD-009 (Automated Discovery Connectors)
• Provenance: LEARN-MAP-001 · Generation standard: LEARN-STD-002
• Next: LEARN-LSN-049