slug: agent-diff
provider: Agent Diff
generated_by: planning/capability-mapping/scripts/classify_capabilities.py
model: claude-opus-5
frame:
- Software & Technology
min_confidence: 0.7
capability_model:
  source: https://github.com/vincentmakes/turbo-ea-capabilities
  license: CC-BY-4.0
  attribution: Turbo EA Capabilities by Vincent Verdet — Turbo EA, https://github.com/vincentmakes/turbo-ea-capabilities,
    CC BY 4.0
  notice: NOTICE
edge_count: 2
edges:
- tag: Diffs
  spec_file: agent-diff-diffs-api-openapi.yml
  capability_id: BC-4200.50
  capability_id_l1: BC-4200
  capability_name: Software Quality Engineering
  confidence: 0.7
  evidence: Agents interact with these sandboxes to produce deterministic state-change diffs without side
    effects, rate limits, or real API calls
  reason: Diff listing is the assertion/verification output used when testing AI agents against faked
    third-party APIs — a test-tooling surface, best read as Software Quality Engineering. Not the full
    test-management discipline, hence 0.7.
- tag: Sandboxes
  spec_file: agent-diff-sandboxes-api-openapi.yml
  capability_id: BC-4210.20
  capability_id_l1: BC-4210
  capability_name: Environment Management
  confidence: 0.7
  evidence: POST /sandboxes create-sandbox ... creates isolated, ephemeral replicas of third-party APIs
    such as Slack and Linear
  reason: Create/list/get/delete of isolated ephemeral sandboxes is provisioning and teardown of non-production
    test environments, matching Environment Management. Could alternatively be read as quality engineering
    tooling, so confidence held at 0.7.