use gh_workflow::{Event, Expression, Job, Run, Step, Strategy, Use, Workflow, WorkflowDispatch};use serde_json::json; use crate::tasks::workflows::{ runners::{self, Platform}, steps::{self, FluentBuilder as _, NamedJob, named, setup_cargo_config}, vars::{self, WorkflowInput},}; pub(crate) fn run_agent_evals() -> Workflow { let agent_evals = agent_evals(); let model_name = WorkflowInput::string("model_name", None); named::workflow() .on(Event::default().workflow_dispatch( WorkflowDispatch::default().add_input(model_name.name, model_name.input()), )) .concurrency(vars::one_workflow_per_non_main_branch()) .add_env(("CARGO_TERM_COLOR", "always")) .add_env(("CARGO_INCREMENTAL", 0)) .add_env(("RUST_BACKTRACE", 1)) .add_env(("ANTHROPIC_API_KEY", vars::ANTHROPIC_API_KEY)) .add_env(("OPENAI_API_KEY", vars::OPENAI_API_KEY)) .add_env(("GOOGLE_AI_API_KEY", vars::GOOGLE_AI_API_KEY)) .add_env(("GOOGLE_CLOUD_PROJECT", vars::GOOGLE_CLOUD_PROJECT)) .add_env(("ZED_CLIENT_CHECKSUM_SEED", vars::ZED_CLIENT_CHECKSUM_SEED)) .add_env(("ZED_EVAL_TELEMETRY", 1)) .add_env(("MODEL_NAME", model_name.to_string())) .add_job(agent_evals.name, agent_evals.job)} pub(crate) fn run_unit_evals() -> Workflow { let model_name = WorkflowInput::string("model_name", None); let commit_sha = WorkflowInput::string("commit_sha", None); let unit_evals = named::job(unit_evals(Some(&commit_sha))); named::workflow() .name("run_unit_evals") .on(Event::default().workflow_dispatch( WorkflowDispatch::default() .add_input(model_name.name, model_name.input()) .add_input(commit_sha.name, commit_sha.input()), )) .concurrency(vars::allow_concurrent_runs()) .add_env(("CARGO_TERM_COLOR", "always")) .add_env(("CARGO_INCREMENTAL", 0)) .add_env(("RUST_BACKTRACE", 1)) .add_env(("ZED_CLIENT_CHECKSUM_SEED", vars::ZED_CLIENT_CHECKSUM_SEED)) .add_env(("ZED_EVAL_TELEMETRY", 1)) .add_env(("MODEL_NAME", model_name.to_string())) .add_job(unit_evals.name, unit_evals.job)} fn add_api_keys(step: Step<Run>) -> Step<Run> { step.add_env(("ANTHROPIC_API_KEY", vars::ANTHROPIC_API_KEY)) .add_env(("OPENAI_API_KEY", vars::OPENAI_API_KEY)) .add_env(("GOOGLE_AI_API_KEY", vars::GOOGLE_AI_API_KEY)) .add_env(("GOOGLE_CLOUD_PROJECT", vars::GOOGLE_CLOUD_PROJECT))} fn agent_evals() -> NamedJob { fn run_eval() -> Step<Run> { named::bash( "cargo run --package=eval -- --repetitions=8 --concurrency=1 --model \"${MODEL_NAME}\"", ) } named::job( Job::default() .runs_on(runners::LINUX_DEFAULT) .timeout_minutes(60_u32 * 10) .add_step(steps::checkout_repo()) .add_step(steps::cache_rust_dependencies_namespace()) .map(steps::install_linux_dependencies) .add_step(setup_cargo_config(Platform::Linux)) .add_step(steps::setup_sccache(Platform::Linux)) .add_step(steps::script("cargo build --package=eval")) .add_step(add_api_keys(run_eval())) .add_step(steps::show_sccache_stats(Platform::Linux)) .add_step(steps::cleanup_cargo_config(Platform::Linux)), )} pub(crate) fn run_cron_unit_evals() -> Workflow { let unit_evals = cron_unit_evals(); named::workflow() .name("run_cron_unit_evals") .on(Event::default() // .schedule([ // // GitHub might drop jobs at busy times, so we choose a random time in the middle of the night. // Schedule::default().cron("47 1 * * 2"), // ]) .workflow_dispatch(WorkflowDispatch::default())) .concurrency(vars::one_workflow_per_non_main_branch()) .add_env(("CARGO_TERM_COLOR", "always")) .add_env(("CARGO_INCREMENTAL", 0)) .add_env(("RUST_BACKTRACE", 1)) .add_env(("ZED_CLIENT_CHECKSUM_SEED", vars::ZED_CLIENT_CHECKSUM_SEED)) .add_job(unit_evals.name, unit_evals.job)} fn cron_unit_evals() -> NamedJob { fn send_failure_to_slack() -> Step<Use> { named::uses( "slackapi", "slack-github-action", "b0fa283ad8fea605de13dc3f449259339835fc52", ) .if_condition(Expression::new("${{ failure() }}")) .add_with(("method", "chat.postMessage")) .add_with(("token", vars::SLACK_APP_ZED_UNIT_EVALS_BOT_TOKEN)) .add_with(("payload", indoc::indoc!{r#" channel: C04UDRNNJFQ text: "Unit Evals Failed: https://github.com/zed-industries/zed/actions/runs/${{ github.run_id }}" "#})) } named::job(cron_unit_evals_job().add_step(send_failure_to_slack()))} const UNIT_EVAL_MODELS: &[&str] = &[ "anthropic/claude-sonnet-4-5-latest", "anthropic/claude-opus-4-5-latest", "google/gemini-3-pro", "openai/gpt-5",]; fn cron_unit_evals_job() -> Job { let script_step = add_api_keys(steps::script("./script/run-unit-evals")) .add_env(("ZED_AGENT_MODEL", "${{ matrix.model }}")); Job::default() .runs_on(runners::LINUX_DEFAULT) .strategy(Strategy::default().fail_fast(false).matrix(json!({ "model": UNIT_EVAL_MODELS }))) .add_step(steps::checkout_repo()) .add_step(steps::setup_cargo_config(Platform::Linux)) .add_step(steps::cache_rust_dependencies_namespace()) .map(steps::install_linux_dependencies) .add_step(steps::cargo_install_nextest()) .add_step(steps::clear_target_dir_if_large(Platform::Linux)) .add_step(steps::setup_sccache(Platform::Linux)) .add_step(script_step) .add_step(steps::show_sccache_stats(Platform::Linux)) .add_step(steps::cleanup_cargo_config(Platform::Linux))} fn unit_evals(commit: Option<&WorkflowInput>) -> Job { let script_step = add_api_keys(steps::script("./script/run-unit-evals")); Job::default() .runs_on(runners::LINUX_DEFAULT) .add_step(steps::checkout_repo()) .add_step(steps::setup_cargo_config(Platform::Linux)) .add_step(steps::cache_rust_dependencies_namespace()) .map(steps::install_linux_dependencies) .add_step(steps::cargo_install_nextest()) .add_step(steps::clear_target_dir_if_large(Platform::Linux)) .add_step(steps::setup_sccache(Platform::Linux)) .add_step(match commit { Some(commit) => script_step.add_env(("UNIT_EVAL_COMMIT", commit)), None => script_step, }) .add_step(steps::show_sccache_stats(Platform::Linux)) .add_step(steps::cleanup_cargo_config(Platform::Linux))}
use gh_workflow::{Event, Expression, Job, Run, Step, Strategy, Use, Workflow, WorkflowDispatch};
use serde_json::json;
use crate::tasks::workflows::{
runners::{self, Platform},
steps::{self, FluentBuilder as _, NamedJob, named, setup_cargo_config},
vars::{self, WorkflowInput},
};
pub(crate) fn run_agent_evals() -> Workflow {
let agent_evals = agent_evals();
let model_name = WorkflowInput::string("model_name", None);
named::workflow()
.on(Event::default().workflow_dispatch(
WorkflowDispatch::default().add_input(model_name.name, model_name.input()),
))
.concurrency(vars::one_workflow_per_non_main_branch())
.add_env(("CARGO_TERM_COLOR", "always"))
.add_env(("CARGO_INCREMENTAL", 0))
.add_env(("RUST_BACKTRACE", 1))
.add_env(("ANTHROPIC_API_KEY", vars::ANTHROPIC_API_KEY))
.add_env(("OPENAI_API_KEY", vars::OPENAI_API_KEY))
.add_env(("GOOGLE_AI_API_KEY", vars::GOOGLE_AI_API_KEY))
.add_env(("GOOGLE_CLOUD_PROJECT", vars::GOOGLE_CLOUD_PROJECT))
.add_env(("ZED_CLIENT_CHECKSUM_SEED", vars::ZED_CLIENT_CHECKSUM_SEED))
.add_env(("ZED_EVAL_TELEMETRY", 1))
.add_env(("MODEL_NAME", model_name.to_string()))
.add_job(agent_evals.name, agent_evals.job)
}
pub(crate) fn run_unit_evals() -> Workflow {
let model_name = WorkflowInput::string("model_name", None);
let commit_sha = WorkflowInput::string("commit_sha", None);
let unit_evals = named::job(unit_evals(Some(&commit_sha)));
named::workflow()
.name("run_unit_evals")
.on(Event::default().workflow_dispatch(
WorkflowDispatch::default()
.add_input(model_name.name, model_name.input())
.add_input(commit_sha.name, commit_sha.input()),
))
.concurrency(vars::allow_concurrent_runs())
.add_env(("CARGO_TERM_COLOR", "always"))
.add_env(("CARGO_INCREMENTAL", 0))
.add_env(("RUST_BACKTRACE", 1))
.add_env(("ZED_CLIENT_CHECKSUM_SEED", vars::ZED_CLIENT_CHECKSUM_SEED))
.add_env(("ZED_EVAL_TELEMETRY", 1))
.add_env(("MODEL_NAME", model_name.to_string()))
.add_job(unit_evals.name, unit_evals.job)
}
fn add_api_keys(step: Step<Run>) -> Step<Run> {
step.add_env(("ANTHROPIC_API_KEY", vars::ANTHROPIC_API_KEY))
.add_env(("OPENAI_API_KEY", vars::OPENAI_API_KEY))
.add_env(("GOOGLE_AI_API_KEY", vars::GOOGLE_AI_API_KEY))
.add_env(("GOOGLE_CLOUD_PROJECT", vars::GOOGLE_CLOUD_PROJECT))
}
fn agent_evals() -> NamedJob {
fn run_eval() -> Step<Run> {
named::bash(
"cargo run --package=eval -- --repetitions=8 --concurrency=1 --model \"${MODEL_NAME}\"",
)
}
named::job(
Job::default()
.runs_on(runners::LINUX_DEFAULT)
.timeout_minutes(60_u32 * 10)
.add_step(steps::checkout_repo())
.add_step(steps::cache_rust_dependencies_namespace())
.map(steps::install_linux_dependencies)
.add_step(setup_cargo_config(Platform::Linux))
.add_step(steps::setup_sccache(Platform::Linux))
.add_step(steps::script("cargo build --package=eval"))
.add_step(add_api_keys(run_eval()))
.add_step(steps::show_sccache_stats(Platform::Linux))
.add_step(steps::cleanup_cargo_config(Platform::Linux)),
)
}
pub(crate) fn run_cron_unit_evals() -> Workflow {
let unit_evals = cron_unit_evals();
named::workflow()
.name("run_cron_unit_evals")
.on(Event::default()
// .schedule([
// // GitHub might drop jobs at busy times, so we choose a random time in the middle of the night.
// Schedule::default().cron("47 1 * * 2"),
// ])
.workflow_dispatch(WorkflowDispatch::default()))
.concurrency(vars::one_workflow_per_non_main_branch())
.add_env(("CARGO_TERM_COLOR", "always"))
.add_env(("CARGO_INCREMENTAL", 0))
.add_env(("RUST_BACKTRACE", 1))
.add_env(("ZED_CLIENT_CHECKSUM_SEED", vars::ZED_CLIENT_CHECKSUM_SEED))
.add_job(unit_evals.name, unit_evals.job)
}
fn cron_unit_evals() -> NamedJob {
fn send_failure_to_slack() -> Step<Use> {
named::uses(
"slackapi",
"slack-github-action",
"b0fa283ad8fea605de13dc3f449259339835fc52",
)
.if_condition(Expression::new("${{ failure() }}"))
.add_with(("method", "chat.postMessage"))
.add_with(("token", vars::SLACK_APP_ZED_UNIT_EVALS_BOT_TOKEN))
.add_with(("payload", indoc::indoc!{r#"
channel: C04UDRNNJFQ
text: "Unit Evals Failed: https://github.com/zed-industries/zed/actions/runs/${{ github.run_id }}"
"#}))
}
named::job(cron_unit_evals_job().add_step(send_failure_to_slack()))
}
const UNIT_EVAL_MODELS: &[&str] = &[
"anthropic/claude-sonnet-4-5-latest",
"anthropic/claude-opus-4-5-latest",
"google/gemini-3-pro",
"openai/gpt-5",
];
fn cron_unit_evals_job() -> Job {
let script_step = add_api_keys(steps::script("./script/run-unit-evals"))
.add_env(("ZED_AGENT_MODEL", "${{ matrix.model }}"));
Job::default()
.runs_on(runners::LINUX_DEFAULT)
.strategy(Strategy::default().fail_fast(false).matrix(json!({
"model": UNIT_EVAL_MODELS
})))
.add_step(steps::checkout_repo())
.add_step(steps::setup_cargo_config(Platform::Linux))
.add_step(steps::cache_rust_dependencies_namespace())
.map(steps::install_linux_dependencies)
.add_step(steps::cargo_install_nextest())
.add_step(steps::clear_target_dir_if_large(Platform::Linux))
.add_step(steps::setup_sccache(Platform::Linux))
.add_step(script_step)
.add_step(steps::show_sccache_stats(Platform::Linux))
.add_step(steps::cleanup_cargo_config(Platform::Linux))
}
fn unit_evals(commit: Option<&WorkflowInput>) -> Job {
let script_step = add_api_keys(steps::script("./script/run-unit-evals"));
Job::default()
.runs_on(runners::LINUX_DEFAULT)
.add_step(steps::checkout_repo())
.add_step(steps::setup_cargo_config(Platform::Linux))
.add_step(steps::cache_rust_dependencies_namespace())
.map(steps::install_linux_dependencies)
.add_step(steps::cargo_install_nextest())
.add_step(steps::clear_target_dir_if_large(Platform::Linux))
.add_step(steps::setup_sccache(Platform::Linux))
.add_step(match commit {
Some(commit) => script_step.add_env(("UNIT_EVAL_COMMIT", commit)),
None => script_step,
})
.add_step(steps::show_sccache_stats(Platform::Linux))
.add_step(steps::cleanup_cargo_config(Platform::Linux))
}