Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.gildea.ai/llms.txt

Use this file to discover all available pages before exploring further.

Gildea uses progressive disclosure to keep responses efficient by default while allowing you to request more detail when needed.

Three levels of detail

Level 1: Signal card (list endpoints)

The lightest view — metadata and a text preview:
GET /v1/signals?entity=NVIDIA
Returns decomposition_text (thesis or summary text), themes, entities, and verified_text_unit_count. No individual text units.

Level 2: Full decomposition (detail endpoint)

The complete verified breakdown with all text units:
GET /v1/signals/0001f3a7b9c8d4e5f6a7b8c9d0e1f2a3b4c5d6e7
Returns the full decomposition tree — every verified sentence and claim with compact verification metadata.

Level 3: With evidence (include=evidence)

Full decomposition plus source evidence snippets:
GET /v1/signals/0001f3a7b9c8d4e5f6a7b8c9d0e1f2a3b4c5d6e7?include=evidence
Each text unit includes its evidence block with a 100-character preview prefix, truncation flag, and full character counts.

Verification detail

Orthogonal to the include parameter, verification_detail controls how much verification metadata is returned:
ValueVerification fields
(default)verdict, path, human_reviewed, scoring_mode, primary_score
full+ scoring_method, components_ran, all scores, thresholds, reason_codes
Combine them:
GET /v1/signals/0001f3a7b9c8d4e5f6a7b8c9d0e1f2a3b4c5d6e7?include=evidence&verification_detail=full

Why progressive disclosure?

  • Bandwidth — signal cards are ~500 bytes; full decomposition with evidence can be 10KB+
  • Cost — fewer bytes transferred = faster responses
  • Clarity — most use cases only need the signal card for browsing