vinzenz/hyperhive
vinzenz/hyperhive
Watch
0
0
Fork
You've already forked hyperhive
 0
Code Activity

swarmctl shares the create-agent, inference and permissions shapes with the controller

Browse source
This commit is contained in:
atlas 2026-10-19 17:57:21 +02:00
commit 0831b85b73
26 changed files with 675 additions and 434 deletions
Download patch file Download diff file Expand all files Collapse all files

69
swarm-controller/src/agent_limits.rs
View file

@ -15,10 +15,9 @@ use axum::Json;
use axum::extract::{Path, State};
use axum::http::StatusCode;
use problem_details::ProblemDetails;
use serde::{Deserialize, Serialize};
use swarm_controller_api::limits::Caps;
use swarm_queue_client::agent_limits;
use swarm_wire::agent_limits::AgentCaps;
use utoipa::ToSchema;
use crate::routes::{AppState, error_problem, wanted_error_status};
use crate::schedules::{live, queue, read_or_default, refused, write_revision};
@ -27,45 +26,19 @@ use crate::schedules::{live, queue, read_or_default, refused, write_revision};
/// `wanted::MAX_ATTEMPTS`.
const MAX_ATTEMPTS: usize = 5;
/// One agent's caps. A field left out is not capped by the swarm.
///
/// Each value is stored as given, and only in a form systemd applies: a value
/// systemd ignores ("Invalid …, ignoring") would leave the agent uncapped,
/// since a swarm cap replaces the hive's answer.
#[derive(Debug, Default, Clone, PartialEq, Eq, Serialize, Deserialize, ToSchema)]
#[serde(deny_unknown_fields)]
pub struct Caps {
/// systemd `CPUQuota=`: digits, optionally `.` and one or two digits, then
/// `%`; above zero and at most `21474836.47%`. `"100%"` is one core.
#[serde(default, skip_serializing_if = "Option::is_none")]
#[schema(example = "400%")]
cpu_quota: Option<String>,
/// systemd `MemoryMax=`: `infinity`; a percentage of physical memory
/// written as for `cpu_quota`, above zero and at most `100%`; or a size:
/// digits, optionally `.` and digits, then nothing, `B` or one of
/// `K`/`M`/`G`/`T` (1024-based, upper case), at least one byte and below
/// 2^64 - 1 bytes.
#[serde(default, skip_serializing_if = "Option::is_none")]
#[schema(example = "8G")]
memory_max: Option<String>,
/// The caps to store, or why they are refused.
pub(crate) fn into_stored(caps: &Caps) -> Result<AgentCaps, String> {
Ok(AgentCaps {
cpu_quota: caps.cpu_quota.as_deref().map(cpu_quota).transpose()?,
memory_max: caps.memory_max.as_deref().map(memory_max).transpose()?,
})
}
impl Caps {
/// The value to store, or why it is refused.
pub(crate) fn into_stored(self) -> Result<AgentCaps, String> {
Ok(AgentCaps {
cpu_quota: self.cpu_quota.as_deref().map(cpu_quota).transpose()?,
memory_max: self.memory_max.as_deref().map(memory_max).transpose()?,
})
}
}
impl From<AgentCaps> for Caps {
fn from(v: AgentCaps) -> Self {
Self {
cpu_quota: v.cpu_quota,
memory_max: v.memory_max,
}
/// The stored caps as the API shows them.
fn view(v: AgentCaps) -> Caps {
Caps {
cpu_quota: v.cpu_quota,
memory_max: v.memory_max,
}
}
@ -213,7 +186,7 @@ pub(crate) async fn get_agent_limits(
let (current, _, _) = read_or_default::<AgentCaps>(&store, &agent)
.await
.map_err(|e| queue_problem(&e))?;
Ok(Json(current.into()))
Ok(Json(view(current)))
}
/// Set the CPU and memory caps the swarm declares for one agent.
@ -244,9 +217,7 @@ pub(crate) async fn put_agent_limits(
Json(req): Json<Caps>,
) -> Result<Json<Caps>, ProblemDetails> {
let agent = agent_name(&name).map_err(refused)?;
let value = req
.into_stored()
.map_err(|why| error_problem(StatusCode::BAD_REQUEST, &why))?;
let value = into_stored(&req).map_err(|why| error_problem(StatusCode::BAD_REQUEST, &why))?;
let client = queue(&state).map_err(refused)?;
if !live(&state).await?.contains(&agent) {
return Err(error_problem(
@ -255,7 +226,7 @@ pub(crate) async fn put_agent_limits(
));
}
store(&client, &agent, &value).await?;
Ok(Json(value.into()))
Ok(Json(view(value)))
}
/// What [`store`] replaced, or `None` when it wrote nothing because the caps
@ -785,10 +756,7 @@ pub(crate) mod tests {
let nats = TestNats::start();
let client = nats.client().await;
let caps = |json: &str| {
serde_json::from_str::<Caps>(json)
.expect("caps")
.into_stored()
.expect("valid")
super::into_stored(&serde_json::from_str::<Caps>(json).expect("caps")).expect("valid")
};
let first = super::store(&client, "atlas", &caps(CAPPED))
@ -814,10 +782,7 @@ pub(crate) mod tests {
let nats = TestNats::start();
let client = nats.client().await;
let caps = |json: &str| {
serde_json::from_str::<Caps>(json)
.expect("caps")
.into_stored()
.expect("valid")
super::into_stored(&serde_json::from_str::<Caps>(json).expect("caps")).expect("valid")
};
let other = caps(r#"{"cpu_quota":"100%"}"#);

52
swarm-controller/src/agent_permissions.rs
View file

@ -19,7 +19,8 @@ use backon::{ConstantBuilder, Retryable as _};
use futures_util::StreamExt as _;
use hive_types::Permission;
use problem_details::ProblemDetails;
use serde::{Deserialize, Serialize};
use serde::Serialize;
use swarm_controller_api::permissions::Permissions;
use swarm_queue_client::agent_permissions;
use swarm_wire::agent_permissions::{
AgentPermissions, Import, ImportReply, hive_from_import_subject, import_subject_wildcard,
@ -34,28 +35,14 @@ use crate::wanted::WantedWriter;
/// `wanted::MAX_ATTEMPTS`.
const MAX_ATTEMPTS: usize = 5;
/// One agent's permissions. A field left out is not set by the swarm, and the
/// compiled-in default applies; an empty list grants nothing.
#[derive(Debug, Default, Clone, PartialEq, Eq, Serialize, Deserialize, ToSchema)]
#[serde(deny_unknown_fields)]
pub struct Permissions {
/// Permission names, `snake_case`: `messaging`, `meta`, `inbox`,
/// `execution`, `web_tools`, `manage_root_agent`.
#[serde(default, skip_serializing_if = "Option::is_none")]
#[schema(value_type = Option<Vec<String>>, example = json!(["messaging", "meta", "inbox", "execution"]))]
permissions: Option<Vec<Permission>>,
}
impl Permissions {
/// The value to store, or why it is refused.
fn into_stored(self) -> Result<AgentPermissions, String> {
if let Some(perms) = &self.permissions {
unique(perms)?;
}
Ok(AgentPermissions {
permissions: self.permissions,
})
/// The permissions to store, or why they are refused.
fn into_stored(req: Permissions) -> Result<AgentPermissions, String> {
if let Some(perms) = &req.permissions {
unique(perms)?;
}
Ok(AgentPermissions {
permissions: req.permissions,
})
}
/// Refuse a list that names a permission twice.
@ -70,11 +57,10 @@ pub(crate) fn unique(perms: &[Permission]) -> Result<(), String> {
Ok(())
}
impl From<AgentPermissions> for Permissions {
fn from(v: AgentPermissions) -> Self {
Self {
permissions: v.permissions,
}
/// The stored permissions as the API shows them.
fn view(v: AgentPermissions) -> Permissions {
Permissions {
permissions: v.permissions,
}
}
@ -122,7 +108,7 @@ pub(crate) async fn get_agent_permissions(
let (current, _, _) = read_or_default::<AgentPermissions>(&store, &agent)
.await
.map_err(|e| queue_problem(&e))?;
Ok(Json(current.into()))
Ok(Json(view(current)))
}
/// Set the permissions the swarm grants one agent.
@ -153,9 +139,7 @@ pub(crate) async fn put_agent_permissions(
Json(req): Json<Permissions>,
) -> Result<Json<Permissions>, ProblemDetails> {
let agent = agent_name(&name).map_err(refused)?;
let value = req
.into_stored()
.map_err(|why| error_problem(StatusCode::BAD_REQUEST, &why))?;
let value = into_stored(req).map_err(|why| error_problem(StatusCode::BAD_REQUEST, &why))?;
let client = queue(&state).map_err(refused)?;
if !live(&state).await?.contains(&agent) {
return Err(error_problem(
@ -174,7 +158,7 @@ pub(crate) async fn put_agent_permissions(
.map_err(|e| queue_problem(&e))?
{
tracing::info!(%agent, ?value, "set agent permissions");
return Ok(Json(value.into()));
return Ok(Json(view(value)));
}
}
Err(queue_problem(&anyhow::anyhow!(
@ -808,7 +792,7 @@ mod tests {
let create = |name: &str, permissions| {
crate::routes::agents::create_agent(
State(state.clone()),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: name.to_owned(),
hive: "pr1ma".to_owned(),
limits: None,
@ -870,7 +854,7 @@ mod tests {
.expect("destroyed");
let _ = crate::routes::agents::create_agent(
State(state.clone()),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,

7
swarm-controller/src/agents_md_initial.rs
View file

@ -138,8 +138,11 @@ mod tests {
.map(|v| String::from_utf8(v.to_vec()).expect("utf-8"))
}
fn request(name: &str, agents_md: Option<String>) -> crate::routes::agents::CreateAgentRequest {
crate::routes::agents::CreateAgentRequest {
fn request(
name: &str,
agents_md: Option<String>,
) -> swarm_controller_api::agents::CreateAgentRequest {
swarm_controller_api::agents::CreateAgentRequest {
name: name.to_owned(),
hive: "pr1ma".to_owned(),
limits: None,

11
swarm-controller/src/archetypes.rs
View file

@ -15,6 +15,8 @@ use axum::extract::State;
use hive_types::Permission;
use problem_details::ProblemDetails;
use serde::{Deserialize, Serialize};
use swarm_controller_api::inference::InferenceSettings;
use swarm_controller_api::limits::Caps;
use swarm_queue_client::archetypes::{ArchetypeMeta, BUILTIN_SOURCE, SourceRecord};
use swarm_wire::builder::{
ARCHETYPES_DEADLINE, ARCHETYPES_SUBJECT, ArchetypesReply, EvalArchetypes,
@ -39,9 +41,9 @@ struct DeclaredMeta {
#[serde(default)]
permissions: Option<Vec<Permission>>,
#[serde(default)]
limits: Option<agent_limits::Caps>,
limits: Option<Caps>,
#[serde(default)]
inference: Option<inference::InferenceSettings>,
inference: Option<InferenceSettings>,
}
/// `meta` as stored, or why a create would refuse it.
@ -68,12 +70,13 @@ fn check(name: &str, meta: serde_json::Value) -> Result<ArchetypeMeta, String> {
permissions,
limits: declared
.limits
.map(agent_limits::Caps::into_stored)
.as_ref()
.map(agent_limits::into_stored)
.transpose()
.map_err(|e| format!("limits: {e}"))?,
inference: declared
.inference
.map(inference::InferenceSettings::into_stored)
.map(inference::into_stored)
.transpose()
.map_err(|e| format!("inference: {e}"))?,
})

204
swarm-controller/src/inference.rs
View file

@ -22,10 +22,11 @@ use axum::extract::{Path, State};
use axum::http::StatusCode;
use futures_util::StreamExt as _;
use problem_details::ProblemDetails;
use serde::{Deserialize, Serialize};
use swarm_controller_api::inference::{
AgentInferenceView, EffectiveInference, InferenceSettings, MAX_SUBAGENTS,
};
use swarm_queue_client::agent_inference;
use swarm_wire::agent_inference::{AgentInference, Effective, Effort, Source};
use utoipa::ToSchema;
use swarm_wire::agent_inference::{AgentInference, Effective};
use crate::routes::{AppState, error_problem, wanted_error_status};
use crate::schedules::{live, queue, read_or_default, refused, write_revision};
@ -34,40 +35,13 @@ use crate::schedules::{live, queue, read_or_default, refused, write_revision};
/// `wanted::MAX_ATTEMPTS`.
const MAX_ATTEMPTS: usize = 5;
/// The highest `max_subagents` a `PUT` may set. Far more concurrent claude
/// sessions than one agent container has memory for, so a value above it is a
/// typo rather than a cap.
const MAX_SUBAGENTS: u32 = 32;
/// One agent's inference settings. An absent field is not set at swarm level,
/// and the agent keeps its own value for it.
#[derive(Debug, Clone, Default, PartialEq, Eq, Serialize, Deserialize, ToSchema)]
#[serde(deny_unknown_fields)]
pub struct InferenceSettings {
/// The model the agent's turns run on, as claude's `--model` takes it.
#[serde(default, skip_serializing_if = "Option::is_none")]
model: Option<String>,
/// One of `low`, `medium`, `high`, `xhigh`, `max`.
#[serde(default, skip_serializing_if = "Option::is_none")]
#[schema(value_type = Option<String>)]
effort: Option<Effort>,
/// The models the agent may start its subagents on. At least one.
#[serde(default, skip_serializing_if = "Option::is_none")]
subagent_models: Option<Vec<String>>,
/// How many subagents the agent may have running at once, from 0 (no
/// subagents) to [`MAX_SUBAGENTS`]. Absent is no cap.
#[serde(default, skip_serializing_if = "Option::is_none")]
max_subagents: Option<u32>,
}
impl From<AgentInference> for InferenceSettings {
fn from(v: AgentInference) -> Self {
Self {
model: v.model,
effort: v.effort,
subagent_models: v.subagent_models,
max_subagents: v.max_subagents,
}
/// The stored settings as the API shows them.
fn settings_view(v: AgentInference) -> InferenceSettings {
InferenceSettings {
model: v.model,
effort: v.effort,
subagent_models: v.subagent_models,
max_subagents: v.max_subagents,
}
}
@ -86,76 +60,47 @@ pub(crate) fn check_model(name: &str, what: &str) -> Result<(), String> {
Ok(())
}
impl InferenceSettings {
/// The value to store, or why it is refused.
pub(crate) fn into_stored(self) -> Result<AgentInference, String> {
if let Some(model) = &self.model {
check_model(model, "model")?;
}
if let Some(models) = &self.subagent_models {
if models.is_empty() {
return Err(
"subagent_models is empty; leave it out to keep the agent's own list"
.to_owned(),
);
}
let mut seen = BTreeSet::new();
for model in models {
check_model(model, "subagent model")?;
if !seen.insert(model) {
return Err(format!("subagent model {model:?} is listed twice"));
}
}
}
if let Some(max) = self.max_subagents
&& max > MAX_SUBAGENTS
{
return Err(format!(
"max_subagents is {max}; it is from 0 (no subagents) to {MAX_SUBAGENTS}, or left out for no cap"
));
}
Ok(AgentInference {
model: self.model,
effort: self.effort,
subagent_models: self.subagent_models,
max_subagents: self.max_subagents,
})
/// The settings to store, or why they are refused.
pub(crate) fn into_stored(settings: InferenceSettings) -> Result<AgentInference, String> {
if let Some(model) = &settings.model {
check_model(model, "model")?;
}
}
/// What `GET` answers for one agent.
#[derive(Debug, Serialize, ToSchema)]
pub struct AgentInferenceView {
/// The settings the swarm sets; `null` when it has none for the agent (no
/// key, or a cleared one), `{}` for a stored setting with no fields.
configured: Option<InferenceSettings>,
/// What the agent reported running on in its latest turn-state header;
/// `null` when this daemon has received none with these fields since it
/// started.
effective: Option<EffectiveInference>,
}
/// The model and effort an agent runs on, and where each comes from.
#[derive(Debug, PartialEq, Eq, Serialize, ToSchema)]
pub struct EffectiveInference {
model: String,
effort: String,
/// `swarm`, `agent_nix`, `local_file` or `default`.
#[schema(value_type = String)]
model_source: Source,
/// As `model_source`.
#[schema(value_type = String)]
effort_source: Source,
}
impl From<Effective> for EffectiveInference {
fn from(v: Effective) -> Self {
Self {
model: v.model,
effort: v.effort,
model_source: v.model_source,
effort_source: v.effort_source,
if let Some(models) = &settings.subagent_models {
if models.is_empty() {
return Err(
"subagent_models is empty; leave it out to keep the agent's own list".to_owned(),
);
}
let mut seen = BTreeSet::new();
for model in models {
check_model(model, "subagent model")?;
if !seen.insert(model) {
return Err(format!("subagent model {model:?} is listed twice"));
}
}
}
if let Some(max) = settings.max_subagents
&& max > MAX_SUBAGENTS
{
return Err(format!(
"max_subagents is {max}; it is from 0 (no subagents) to {MAX_SUBAGENTS}, or left out for no cap"
));
}
Ok(AgentInference {
model: settings.model,
effort: settings.effort,
subagent_models: settings.subagent_models,
max_subagents: settings.max_subagents,
})
}
/// The agent's reported values as the API shows them.
fn effective_view(v: Effective) -> EffectiveInference {
EffectiveInference {
model: v.model,
effort: v.effort,
model_source: v.model_source,
effort_source: v.effort_source,
}
}
@ -268,12 +213,12 @@ pub(crate) async fn get_agent_inference(
.await
.map_err(|e| queue_problem(&e))?;
Ok(Json(AgentInferenceView {
configured: current.map(Into::into),
configured: current.map(settings_view),
effective: state
.reported_inference
.as_ref()
.and_then(|r| r.get(&agent))
.map(Into::into),
.map(effective_view),
}))
}
@ -304,9 +249,7 @@ pub(crate) async fn put_agent_inference(
Json(req): Json<InferenceSettings>,
) -> Result<Json<InferenceSettings>, ProblemDetails> {
let agent = agent_name(&name).map_err(refused)?;
let value = req
.into_stored()
.map_err(|why| error_problem(StatusCode::BAD_REQUEST, &why))?;
let value = into_stored(req).map_err(|why| error_problem(StatusCode::BAD_REQUEST, &why))?;
let client = queue(&state).map_err(refused)?;
if !live(&state).await?.contains(&agent) {
return Err(error_problem(
@ -315,7 +258,7 @@ pub(crate) async fn put_agent_inference(
));
}
store(&client, &agent, &value).await?;
Ok(Json(value.into()))
Ok(Json(settings_view(value)))
}
/// What [`store`] replaced, or `None` when it wrote nothing because the
@ -413,6 +356,7 @@ pub(crate) async fn delete_agent_inference(
#[cfg(test)]
mod tests {
use super::*;
use swarm_controller_api::inference::{Effort, Source};
fn settings(json: &str) -> serde_json::Result<InferenceSettings> {
serde_json::from_str(json)
@ -420,10 +364,10 @@ mod tests {
#[test]
fn a_full_put_is_stored_as_sent() {
let stored = settings(r#"{"model":"opus","effort":"high","subagent_models":["haiku"]}"#)
.unwrap()
.into_stored()
.unwrap();
let stored = into_stored(
settings(r#"{"model":"opus","effort":"high","subagent_models":["haiku"]}"#).unwrap(),
)
.unwrap();
assert_eq!(
stored,
AgentInference {
@ -434,7 +378,7 @@ mod tests {
}
);
assert_eq!(
serde_json::to_value(InferenceSettings::from(stored)).unwrap(),
serde_json::to_value(settings_view(stored)).unwrap(),
serde_json::json!({"model": "opus", "effort": "high", "subagent_models": ["haiku"]})
);
}
@ -442,10 +386,7 @@ mod tests {
/// Each field is optional, and one left out stays out of the stored value.
#[test]
fn a_partial_put_leaves_the_other_fields_unset() {
let stored = settings(r#"{"effort":"max"}"#)
.unwrap()
.into_stored()
.unwrap();
let stored = into_stored(settings(r#"{"effort":"max"}"#).unwrap()).unwrap();
assert_eq!(
stored,
AgentInference {
@ -516,7 +457,7 @@ mod tests {
let view = |effective: Option<Effective>| {
serde_json::to_value(AgentInferenceView {
configured: Some(InferenceSettings::default()),
effective: effective.map(Into::into),
effective: effective.map(effective_view),
})
.unwrap()
};
@ -553,28 +494,21 @@ mod tests {
r#"{"model":"-x"}"#,
r#"{"subagent_models":["haiku","--print"]}"#,
] {
assert!(settings(json).unwrap().into_stored().is_err(), "{json}");
assert!(into_stored(settings(json).unwrap()).is_err(), "{json}");
}
}
#[test]
fn max_subagents_outside_its_range_is_refused() {
assert!(
settings(r#"{"max_subagents":33}"#)
.unwrap()
.into_stored()
.is_err()
);
assert!(into_stored(settings(r#"{"max_subagents":33}"#).unwrap()).is_err());
assert!(settings(r#"{"max_subagents":-1}"#).is_err());
}
#[test]
fn max_subagents_in_range_is_stored() {
for max in [0, 1, 3, MAX_SUBAGENTS] {
let stored = settings(&format!(r#"{{"max_subagents":{max}}}"#))
.unwrap()
.into_stored()
.unwrap();
let stored =
into_stored(settings(&format!(r#"{{"max_subagents":{max}}}"#)).unwrap()).unwrap();
assert_eq!(stored.max_subagents, Some(max));
}
}
@ -583,7 +517,7 @@ mod tests {
#[test]
fn a_dash_inside_a_model_name_is_accepted() {
let json = r#"{"model":"claude-opus-4-8","subagent_models":["claude-haiku-4-5"]}"#;
assert!(settings(json).unwrap().into_stored().is_ok());
assert!(into_stored(settings(json).unwrap()).is_ok());
}
/// What `get_agent_inference` reads with: an absent or deleted key is
@ -638,7 +572,7 @@ mod tests {
let bucket = agent_inference::open_or_create(&client)
.await
.expect("bucket");
let value = |json: &str| settings(json).unwrap().into_stored().unwrap();
let value = |json: &str| into_stored(settings(json).unwrap()).unwrap();
let first = store(&client, "atlas", &value(r#"{"model":"opus"}"#))
.await
@ -666,7 +600,7 @@ mod tests {
let bucket = agent_inference::open_or_create(&client)
.await
.expect("bucket");
let value = |json: &str| settings(json).unwrap().into_stored().unwrap();
let value = |json: &str| into_stored(settings(json).unwrap()).unwrap();
let made = store(&client, "atlas", &value(r#"{"model":"opus"}"#))
.await

10
swarm-controller/src/jobq.rs
View file

@ -1681,7 +1681,7 @@ mod tests {
let (state, sched) = state_with_roster();
let _queued = crate::routes::agents::create_agent(
axum::extract::State(state),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,
@ -1804,7 +1804,7 @@ mod tests {
let (state, sched) = state_with_roster();
let _queued = crate::routes::agents::create_agent(
axum::extract::State(state),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,
@ -1856,7 +1856,7 @@ mod tests {
let (state, sched) = state_with_roster();
let _queued = crate::routes::agents::create_agent(
axum::extract::State(state),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,
@ -1916,7 +1916,7 @@ mod tests {
let (state, sched) = state_with_roster();
let _queued = crate::routes::agents::create_agent(
axum::extract::State(state),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,
@ -2123,7 +2123,7 @@ mod tests {
let (state, sched) = state_with_roster();
let _queued = crate::routes::agents::create_agent(
axum::extract::State(state),
axum::Json(crate::routes::agents::CreateAgentRequest {
axum::Json(swarm_controller_api::agents::CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,

111
swarm-controller/src/routes/agents.rs
View file

@ -9,7 +9,7 @@ use axum::{
extract::{Path, State},
};
use serde::{Deserialize, Serialize};
use swarm_controller_api::agents::CreateAgentResponse;
use swarm_controller_api::agents::{CreateAgentRequest, CreateAgentResponse};
use utoipa::ToSchema;
use crate::jobq::{SwarmNodeKind, SwarmResourceKind, declare_agent_job};
@ -467,63 +467,6 @@ pub(crate) async fn get_agents_status(
}
}
/// Body of `POST /api/agents` — the agent name to create, and the hive the
/// creation is aimed at. The repo name inside `forge::CONFIG_ORG` is the
/// same string as `name`: one repo per agent, named after it, same
/// convention `hive-c0re::forge` already uses for its own single-hive
/// `CreateRepo` path.
///
/// `hive` is an **address**, not an attribute of the agent: it is where a
/// deploy message goes over the queue. Nothing here writes it into the
/// agent's config — a config naming its own hive would be a second
/// statement of where the agent lives, free to drift from the queue that
/// actually delivers to it.
///
/// It is required rather than optional because it is only knowable here,
/// at creation, from the operator making the choice. An optional field
/// would have been the friendlier migration and would have left every
/// creation in the state this endpoint exists to avoid — one nothing can
/// address.
///
/// Validated here and carried no further: none of the nodes this endpoint
/// queues talks to a hive, so none of them needs the address. It reaches
/// its consumer when the node that *sends* a deploy message exists, and
/// that node takes it from this field. Settling the request shape now is
/// the point — it is the breaking half, and doing it once is cheaper for
/// every caller than doing it again later.
#[derive(Clone, Debug, Deserialize, ToSchema)]
pub(crate) struct CreateAgentRequest {
pub(crate) name: String,
pub(crate) hive: String,
/// CPU and memory caps to set for the agent, as `PUT
/// /api/agents/{name}/limits` takes them. Absent leaves any caps already
/// stored for the name alone.
#[serde(default)]
pub(crate) limits: Option<agent_limits::Caps>,
/// Inference settings to set for the agent, as `PUT
/// /api/agents/{name}/inference` takes them. Absent as `limits`.
#[serde(default)]
pub(crate) inference: Option<inference::InferenceSettings>,
/// The agent's permissions, stored before the job is queued. Left out, the
/// agent holds the default set.
#[serde(default)]
#[schema(value_type = Option<Vec<String>>)]
pub(crate) permissions: Option<Vec<hive_types::Permission>>,
/// The archetype to build the agent from, one `GET /api/archetypes`
/// lists. The agent then gets no `agent-configs/<name>` repo: its hive
/// builds it from the archetype's module. Absent is a custom agent.
#[serde(default)]
#[schema(value_type = Option<Object>, example = json!({"source": "hyperhive", "name": "reviewer"}))]
pub(crate) archetype: Option<swarm_queue_client::agent_archetype::ArchetypeRef>,
/// The text the agent's `AGENTS.md` starts with, at most
/// `swarm_queue_client::agents_md_initial::MAX_BYTES` bytes. The agent's
/// harness writes it on its first start, and only when the agent has no
/// `AGENTS.md` yet: an existing file is never replaced. Absent leaves the
/// file to the harness.
#[serde(default)]
pub(crate) agents_md: Option<String>,
}
/// Queue the whole agent-creation job graph for `name` — two independent
/// roots, `CreateIdentity` and `CreateForgeUser`, since an authelia
/// subject and a forge account need nothing from each other. `CreateRepo`
@ -739,10 +682,8 @@ impl CreateSettings {
}
}
Ok(Self {
limits: limits.map(agent_limits::Caps::into_stored).transpose()?,
inference: inference
.map(inference::InferenceSettings::into_stored)
.transpose()?,
limits: limits.as_ref().map(agent_limits::into_stored).transpose()?,
inference: inference.map(inference::into_stored).transpose()?,
permissions: permissions.map(agent_permissions::at_create).transpose()?,
archetype,
agents_md,
@ -1256,7 +1197,7 @@ mod tests {
let err = super::create_agent(
axum::extract::State(state),
axum::Json(super::CreateAgentRequest {
axum::Json(CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1maa".to_owned(),
limits: None,
@ -1299,7 +1240,7 @@ mod tests {
// accept arm is actually for is the queue count below.
let _queued = super::create_agent(
axum::extract::State(state),
axum::Json(super::CreateAgentRequest {
axum::Json(CreateAgentRequest {
name: "atlas".to_owned(),
hive: "pr1ma".to_owned(),
limits: None,
@ -1325,18 +1266,18 @@ mod tests {
/// existed, and each field is checked by the same rules as its `PUT`.
#[test]
fn a_creation_request_takes_optional_settings() {
let plain: super::CreateAgentRequest =
let plain: CreateAgentRequest =
serde_json::from_str(r#"{"name":"atlas","hive":"pr1ma"}"#).expect("plain body");
assert!(plain.limits.is_none() && plain.inference.is_none());
let full: super::CreateAgentRequest = serde_json::from_str(
let full: CreateAgentRequest = serde_json::from_str(
r#"{"name":"atlas","hive":"pr1ma","limits":{"cpu_quota":"400%"},
"inference":{"model":"opus","effort":"high"}}"#,
)
.expect("body with settings");
assert!(full.limits.is_some() && full.inference.is_some());
let typo = serde_json::from_str::<super::CreateAgentRequest>(
let typo = serde_json::from_str::<CreateAgentRequest>(
r#"{"name":"atlas","hive":"pr1ma","limits":{"cpu":"400%"}}"#,
);
assert!(typo.is_err(), "an unknown cap field must be refused");
@ -1355,7 +1296,7 @@ mod tests {
for (value, field) in bad {
let (state, sched) = state_with_roster();
let body = format!(r#"{{"name":"atlas","hive":"pr1ma","{field}":{value}}}"#);
let req: super::CreateAgentRequest = serde_json::from_str(&body).expect("body parses");
let req: CreateAgentRequest = serde_json::from_str(&body).expect("body parses");
let err = super::create_agent(axum::extract::State(state), axum::Json(req))
.await
.expect_err("a bad setting must be refused");
@ -1373,7 +1314,7 @@ mod tests {
#[tokio::test]
async fn settings_that_cannot_be_stored_refuse_the_creation() {
let (state, sched) = state_with_roster();
let req: super::CreateAgentRequest = serde_json::from_str(
let req: CreateAgentRequest = serde_json::from_str(
r#"{"name":"atlas","hive":"pr1ma","limits":{"cpu_quota":"400%"}}"#,
)
.expect("body parses");
@ -1394,7 +1335,7 @@ mod tests {
let nats = crate::test_nats::TestNats::start();
let client = nats.client().await;
let state = crate::agent_limits::tests::state(&client);
let req: super::CreateAgentRequest = serde_json::from_str(
let req: CreateAgentRequest = serde_json::from_str(
r#"{"name":"atlas","hive":"pr1ma","limits":{"cpu_quota":"400%"},
"inference":{"model":"opus","max_subagents":3}}"#,
)
@ -1449,16 +1390,18 @@ mod tests {
.await
.expect("a bucket that refuses any value");
let caps = |json: &str| {
serde_json::from_str::<crate::agent_limits::Caps>(json)
.expect("caps")
.into_stored()
.expect("valid")
crate::agent_limits::into_stored(
&serde_json::from_str::<swarm_controller_api::limits::Caps>(json).expect("caps"),
)
.expect("valid")
};
let inference =
serde_json::from_str::<crate::inference::InferenceSettings>(r#"{"model":"opus"}"#)
.expect("settings")
.into_stored()
.expect("valid");
let inference = crate::inference::into_stored(
serde_json::from_str::<swarm_controller_api::inference::InferenceSettings>(
r#"{"model":"opus"}"#,
)
.expect("settings"),
)
.expect("valid");
crate::agent_limits::store(&client, "atlas", &caps(r#"{"cpu_quota":"100%"}"#))
.await
.expect("an earlier cap");
@ -1515,7 +1458,7 @@ mod tests {
.expect("an earlier grant");
for agent in ["atlas", "iris"] {
let req: super::CreateAgentRequest = serde_json::from_str(&format!(
let req: CreateAgentRequest = serde_json::from_str(&format!(
r#"{{"name":"{agent}","hive":"pr1ma","permissions":["web_tools"],
"inference":{{"model":"opus"}}}}"#
))
@ -1560,7 +1503,7 @@ mod tests {
let nats = crate::test_nats::TestNats::start();
let client = nats.client().await;
let state = crate::agent_limits::tests::state(&client);
let req: super::CreateAgentRequest = serde_json::from_str(
let req: CreateAgentRequest = serde_json::from_str(
r#"{"name":"atlas","hive":"pr1ma","limits":{"cpu_quota":"400%"},
"inference":{"model":"-opus"}}"#,
)
@ -1602,7 +1545,7 @@ mod tests {
) -> Result<(), problem_details::ProblemDetails> {
super::create_agent(
axum::extract::State(state.clone()),
axum::Json(super::CreateAgentRequest {
axum::Json(CreateAgentRequest {
name: name.to_owned(),
hive: hive.to_owned(),
limits: None,
@ -2243,7 +2186,7 @@ mod tests {
let client = nats.client().await;
seed_archetypes(&client).await;
let state = crate::agent_limits::tests::state(&client);
let req: super::CreateAgentRequest = serde_json::from_str(
let req: CreateAgentRequest = serde_json::from_str(
r#"{"name":"atlas","hive":"pr1ma",
"archetype":{"source":"hyperhive","name":"reviewer"}}"#,
)
@ -2285,7 +2228,7 @@ mod tests {
r#"{"source":"elsewhere","name":"reviewer"}"#,
r#"{"source":"hyperhive","name":"a.b"}"#,
] {
let req: super::CreateAgentRequest = serde_json::from_str(&format!(
let req: CreateAgentRequest = serde_json::from_str(&format!(
r#"{{"name":"atlas","hive":"pr1ma","permissions":["messaging"],
"archetype":{archetype}}}"#
))