# `mix agent_db.seed`
[🔗](https://github.com/manhvu/agent_db/blob/v0.1.0/lib/mix/tasks/agent_db.seed.ex#L1)

Seeds a deterministic demo dataset for dev and test environments.

    mix agent_db.seed [--prefix URI] [--force] [--clean] [--allow-prod] [--json]

Builds the same baseline on every run: documents under the demo prefix
(default `viking://resources/demo`), one typed memory per taxonomy type,
one session committed into the tree, one demo skill, and two indexed Elixir
sources for the demo project. Re-running converges without duplication.

A non-empty store is refused unless `--force` (merge: missing seed URIs
are created, present ones revised in place, nothing deleted) or `--clean`
(only the seed scope is removed first: the prefix subtree, the `demo-*`
memories, the demo skill, and the demo code prefix; data outside that scope
survives) is given. A refusal writes nothing. `--force` and `--clean`
cannot be combined.

In production (`Mix.env() == :prod`) seeding is refused unless
`--allow-prod` is passed. Rollback is `mix agent_db.read` to inspect,
then removing the prefix subtree and forgetting the `demo-*` memory URIs;
re-seeding afterwards converges.

Reads need no model: keyword search, find, and grep reach seeded content
immediately, while embeddings and summaries regenerate through the existing
job queue.

Options:

  * `--prefix` - demo prefix for documents and the committed session
    (default `viking://resources/demo`).
  * `--force` - merge into a non-empty store instead of refusing it.
  * `--clean` - remove only the seed scope, then seed.
  * `--allow-prod` - allow seeding when `Mix.env() == :prod`.
  * `--json` - print the result as JSON instead of human text.
  * `--no-compile` - do not compile the project before running.

---

*Consult [api-reference.md](api-reference.md) for complete listing*