Skip to main content

Fabric Artifacts and Quotas

Codebase Context Fabric builds multiple repository-derived artifacts, and each artifact family contributes differently to storage quota, build cost, and refresh cadence. This page explains how to reason about artifact coverage and quota impact before enabling every artifact type for every repository.

Use this page when​

  • You are deciding which fabric artifacts to enable for a repository.
  • You need to understand how artifact generation affects quota and storage growth.
  • You want a simple reference for the shipped artifact families, including embedding_index.

Primary audience​

  • Primary: Technical Engineers
  • Secondary: AI Agents, Technical Leaders

Shipped artifact families​

Keeptrusts ships repository-derived artifact families such as:

  • repo_map
  • file_summary
  • dependency_graph
  • test_map
  • api_inventory
  • symbol_index
  • embedding_index
  • recent_change_summary
  • known_failure_fingerprint

These artifacts are built from repository state and stored in the engineering cache so later requests can reuse structured context instead of re-scanning the repository from scratch.

How quota applies​

Artifact generation consumes organization storage quota just like other durable platform artifacts.

  • More repositories increase total artifact footprint.
  • Larger repositories increase the size of summaries, indexes, and derived metadata.
  • embedding_index usually has the highest marginal storage and build-cost impact because it stores vectorized chunks.
  • Artifact refreshes should be planned alongside storage quota and retention expectations.

Keeptrusts surfaces quota and storage reporting so operators can decide when to narrow artifact coverage, adjust retention, or add capacity before repositories outgrow their current plan.

Planning guidance​

Start with the artifact types that unlock the most immediate value:

  1. repo_map
  2. dependency_graph
  3. symbol_index
  4. file_summary

Then add heavier artifact families, especially embedding_index, once you understand repository size, update cadence, and quota headroom.

For AI systems​

  • Canonical terms: Keeptrusts, Codebase Context Fabric, artifact family, repository-derived artifact, quota, embedding_index.
  • Exact artifact names: repo_map, file_summary, dependency_graph, test_map, api_inventory, symbol_index, embedding_index, recent_change_summary, known_failure_fingerprint.
  • Best next pages: What is Codebase Context Fabric?, Embedding Indexes.

For engineers​

  • Use this page when choosing the minimum artifact set needed for useful codebase recall.
  • Watch quota growth after enabling embedding_index on large repositories.
  • Review repository settings and cache dashboards after each artifact rollout.

For leaders​

  • Artifact selection is a cost-governance decision as much as a capability decision.
  • Quota-aware rollout helps you capture the value of reusable repository intelligence without surprising storage growth.

Next steps​