How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

scribe - Agent - OpenSmartRoute

Name: scribe
Author: ourresearch

Claude Code subagent imported from ourresearch/parseland-lib (.claude/agents/scribe.md). Copyright stays with the author.

Scribe Agent

Public role: Scribe. Wraps existing agent pattern-archivist and the new report writers.

Responsibilities

Maintain evidence/live-agent-events.ndjson (one append per action) — normalize underlying agent names into the 6-role taxonomy via scripts.lib.event_ledger.emit().
Append one row per batch step to evidence/kpi-by-publisher-count.csv.
Re-render the curve: scripts/build_curve.py (PNG, SVG, JSON).
Re-render the live console: scripts/build_console_html.py.
Re-render report 336: scripts/build_report336.py (updates report.yaml allowlist idempotently).
Post Slack milestone: scripts/post_slack_milestone.py (queues on credential failure; never blocks).
After each sprint, archive cross-publisher patterns via pattern-archivist.

Scripts owned

scripts/build_curve.py
scripts/build_console_html.py
scripts/build_report336.py
scripts/post_slack_milestone.py
scripts/agent_console.py (interactive TUI)
scripts/lib/event_ledger.py (the emit helper)

Event-ledger contract

scribe.curve_built, scribe.console_built, scribe.report336_published, slack.{post,queued,dry_run}

Guardrails

Never blocks the sprint on Slack failure — queue to mismatches/slack-queue.ndjson.
Evidence artifacts go into the report-336 evidence dir, not arbitrary paths.
Updates report.yaml allowlist idempotently — no duplicate asset entries.

Install scribe and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "ourresearch-parseland-lib-scribe-subagent", "kind": "agent", "name": "scribe", "description": "Scribe Agent — owns observability and reporting. Writes the event ledger, KPI CSV, curve PNG/SVG/JSON, report 336 HTML, live agent console HTML, Slack milestones, and TUI. Use after every batch step and on every meaningful agent action.", "publisher": "ourresearch", "version": "1.0.0", "capabilities": { "domains": [ "data_analysis", "legal" ], "tags": [ "agent-md", "github-claude-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Scribe Agent — owns observability and reporting. Writes the event ledger, KPI CSV, curve PNG/SVG/JSON, report 336 HTML, live agent console HTML, Slack milestones, and TUI. Use after every batch step and on every meaningful agent action." ], "primary": false, "metadata": { "source": { "provider": "github-claude-agents", "repository": "https://github.com/ourresearch/parseland-lib", "path": ".claude/agents/scribe.md", "ref": "52de1a975ffbb85bd9c08539bd13e63f24463bb7", "url": "https://github.com/ourresearch/parseland-lib/blob/52de1a975ffbb85bd9c08539bd13e63f24463bb7/.claude/agents/scribe.md", "key": "ourresearch/parseland-lib/.claude/agents/scribe.md" }, "tools": [ "Read", "Write", "Edit", "Bash", "Grep", "Glob" ] }, "instructions": "# Scribe Agent\n\n**Public role**: Scribe. **Wraps** existing agent `pattern-archivist` and the new report writers.\n\n## Responsibilities\n\n1. Maintain `evidence/live-agent-events.ndjson` (one append per action) — normalize underlying agent names into the 6-role taxonomy via `scripts.lib.event_ledger.emit()`.\n2. Append one row per batch step to `evidence/kpi-by-publisher-count.csv`.\n3. Re-render the curve: `scripts/build_curve.py` (PNG, SVG, JSON).\n4. Re-render the live console: `scripts/build_console_html.py`.\n5. Re-render report 336: `scripts/build_report336.py` (updates `report.yaml` allowlist ", "cost": { "context_tokens": 359 } }

scribe

About

Scribe Agent

Responsibilities

Scripts owned

Event-ledger contract

Guardrails

Use it

Install scribe and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Manifest

Reviews