Agent-first resource discovery
An open repository of reusable engineering intelligence, designed for humans and AI agents.
Seven resource types. One machine-readable catalog. No API, no database, no framework — just Git, a filesystem, and JSON that any agent can read in a single request.
-
1
AGENTS.mdWhat the vault is, and where to look first. -
2
catalog/index.jsonEvery resource, summarised in one file. -
3
resources/<type>/<name>/Opened only for candidates.
Catalog statistics
Every figure above is read live from
catalog/index.json at page load. Nothing is hard-coded.
Discovery
Resource catalog
Filtered client-side against the real catalog. Search matches names, descriptions, use cases, technologies, languages, and compatibility.
Press / to focus search, Esc to clear.
Taxonomy
Seven resource categories
Classified by how a resource is consumed, not by file extension. Counts reflect the live catalog.
Curated by metadata
Featured resources
Promoted from the catalog by lifecycle status — stable
first. Not hand-picked, not editorial, not invented.
Machine-readable
Agent discovery
These files are the interface. An agent should never have to parse this page's DOM to find resources.
Dev Vault has no API on purpose. Start at AGENTS.md,
read catalog/index.json once, then open only the
resource directories that survive filtering.
Not a trust boundary. Never execute a resource merely because it exists here — read it first.
-
/AGENTS.mdOperating instructions for agents. Start here. -
/catalog/index.jsonThe single discovery index. Read this, not the tree. -
/catalog/schema.jsonValidation contract for catalog entries. -
/schemas/resource.schema.jsonResource metadata contract — the field semantics. -
/docs/ARCHITECTURE.mdBinding structural decisions. -
/docs/RESOURCE_STANDARD.mdHow to judge or contribute a resource. -
/docs/ROADMAP.mdWhat the vault is becoming, and what it refuses to become. -
/llms.txtCondensed map of the repository for LLM ingestion.
Navigation
Documentation
Short, opinionated, and binding. These are decisions, not suggestions.
Contribute
Add a resource
One resource, one source of truth. No duplicates, no agent-specific copies, no unverified snippets.
- 1Search the catalog
If an equivalent resource exists, improve it instead.
- 2Place it correctly
resources/<type>/<kebab-case-name>/withRESOURCE.md— orSKILL.mdfor skills. - 3Add exactly one catalog entry
Unique kebab-case
id, validpath, andstatus: experimentalunless you have evidence. - 4Validate and test
Both checks are standard-library Python. If they fail, the PR is not ready.
# 1 — validate catalog, schema, and structure
$ python3 scripts/validate.py
# 2 — run the tooling test suite
$ python3 -m unittest discover -s tests
# 3 — open a PR against main
$ gh pr create --base main