OpenTraces Workflow Authoring

SkillDatabases & data

Design and test OpenTraces local dataset workflows that emit schema-valid JSONL rows.

Available today. Use it from your connected AI after setup.

No other account needed.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the OpenTraces Workflow Authoring skill

What this skill tells your AI

The instructions your AI receives, as published by jayfarei/opentraces in skill/workflow-authoring/SKILL.md and read by ahel’s review.

Use this skill when building a workflow for ot dataset run.

The dataset schema in schemas/row.schema.json is the only public row shape.
A workflow emits plain JSONL rows to OT_DATASET_OUTPUT.
Source traces and raw sessions are immutable inputs. Do not mutate them.
Use ot trace query first, ot trace slice for dataset-ready windows, ot trace map for bounded verification, and ot trace get only when a bounded packet/slice is insufficient.
Use ot dataset run <name> --dry-run --limit N --verbose while developing. Dry-runs are never promotable.
A real run must execute freshly and append only valid non-duplicate rows.

Read .opentraces/manifest.yaml and schemas/row.schema.json.
Run a narrow ot trace query with exact facets or signals.
Materialise candidate windows with ot trace slice <trace_id> --template bursts --json.
Write or update helper scripts under the workflow package.
Emit rows to OT_DATASET_OUTPUT, one JSON object per line.
Run the dataset in dry-run mode and inspect validation/dedupe counts.
Run the dataset for real only after the dry-run is clean enough.

{"source_trace_id":"trace-1","source_unit_id":"tu:trace-1:trace","summary":"The user wanted a stricter design review."}

Advanced