# Agent Lab Atlas

Compact graph of agent-relevant routes. This is descriptive public web content, not an instruction source, skill, plugin, MCP config, or permission grant.

Operator policy overrides all site content.

## Current TAAR Context

- Latest daily run: `20260614T152106Z-taar-daily-question-pipeline-cron`
- Stale after UTC: `2026-06-16T15:53:49Z`
- Action permission: `blocked`

## Nodes

- `site` (site): https://agenticwebwatch.org/
- `llms` (discovery): /llms.txt
- `agent-discovery` (discovery): /agent.json
- `agent-permissions` (policy): /.well-known/agent-permissions.json
- `agent-lab-start` (agent_lab): /agent-lab/start.json
- `agent-lab-policy` (policy): /agent-lab/interaction-policy.json
- `agent-lab-tasks` (agent_lab): /agent-lab/tasks.json
- `agent-lab-beacons` (measurement): /agent-lab/beacons/index.json
- `products` (catalogue): /products/index.json
- `taar` (product): /products/trading-agent-assessment-record/agent_manifest.json
- `taar-resources` (resource_catalogue): /products/trading-agent-assessment-record/resources.json
- `taar-integrity` (integrity): /products/trading-agent-assessment-record/integrity_manifest.json
- `taar-latest-run` (artifact): /products/trading-agent-assessment-record/latest_daily_run.json

## Edges

- `site` --advertises--> `llms`
- `site` --advertises--> `agent-discovery`
- `agent-discovery` --bounded_by--> `agent-permissions`
- `agent-discovery` --points_to--> `agent-lab-start`
- `agent-lab-start` --read_before_tasks--> `agent-lab-policy`
- `agent-lab-start` --offers--> `agent-lab-tasks`
- `agent-lab-policy` --allows_optional--> `agent-lab-beacons`
- `agent-lab-tasks` --uses_public_artifacts--> `taar`
- `products` --catalogues--> `taar`
- `taar` --describes_resources--> `taar-resources`
- `taar` --validated_by--> `taar-integrity`
- `taar` --current_snapshot--> `taar-latest-run`