@oisincoveney/pipeline 1.5.6 → 1.6.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -124,10 +124,17 @@ orchestrator:
124
124
 
125
125
  workflows:
126
126
  default:
127
+ execution:
128
+ fail_fast: true
129
+ max_parallel_nodes: 2
127
130
  nodes:
128
131
  - id: implement
129
132
  kind: agent
130
133
  profile: implementer
134
+ timeout_ms: 300000
135
+ retries:
136
+ max_attempts: 2
137
+ retry_on: [exit_nonzero, gate_failure, timeout]
131
138
  gates:
132
139
  - kind: builtin
133
140
  builtin: test
@@ -189,6 +196,9 @@ runners:
189
196
  - Native subagent strategy is preferred when the selected runner can represent
190
197
  the configured semantics. Otherwise the runtime uses a subprocess boundary.
191
198
  - Parallel DAG batches run concurrently after dependencies and gates pass.
199
+ - Workflow execution can cap parallelism and enable fail-fast batch stopping.
200
+ - Nodes can declare bounded retries, retry reasons, backoff, and execution
201
+ timeouts.
192
202
  - Agent self-reporting is not enough to pass deterministic gates.
193
203
  - JSON Schema gates validate structure only. Use `verdict` and `acceptance`
194
204
  gates to enforce semantic pass/fail and per-criterion coverage.