The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS.
Rust agent-orchestrationai-agentsai-assistantdesktopllmlocal-firstmcppersonal-aiprivacyrustsecond-braintauri
test(agent-harness): route describe_workflow through the workflows pack
Steven Enamakel263a3dfd4f25
diff --git a/tests/agent_harness_e2e.rs b/tests/agent_harness_e2e.rsindex 69847cb91a..200e282f1d 100644--- a/tests/agent_harness_e2e.rs+++ b/tests/agent_harness_e2e.rs@@ -3700,7 +3700,8 @@ async fn agent_installs_a_registry_skill_then_runs_it_inner() {text_completion("Installed the skill."),// The delivery turn: the orchestrator loads the installed skill itself;// running one is its own `run_workflow`.- tool_call_completion(+ packed_tool_call_completion(+ "workflows","describe_workflow",json!({ "workflow_id": REGISTRY_SKILL_ID }),),@@ -3805,7 +3806,9 @@ async fn agent_installs_a_registry_skill_then_runs_it_inner() {);// Usable: the orchestrator loaded the installed skill and its body reached the model.- let described = tool_result_text(&requests, "describe_workflow")+ // `describe_workflow` is a member of the `workflows` pack: the call goes+ // through `use_skill`, and its result is that call's tool result.+ let described = tool_result_text(&requests, "use_skill").unwrap_or_else(|| panic!("no tool result for describe_workflow; requests: {}", dump()));assert!(described.contains(SKILL_BODY_CANARY),