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.
A signal is a structured takeaway drawn from the most strategically relevant parts of a source. A signal may represent a full article, a segment of a podcast, a section of an SEC filing, or any other discrete unit of content worth tracking. Gildea processes signals daily through an automated pipeline, prioritizing what materially changes the strategic picture.
Signal scope
Signal scope is not the same as source scope. A two-hour podcast might yield one signal from a five-minute exchange; a short news article might be captured in its entirety. The title, url, and published_at fields always point to the full source, but the decomposition (thesis, arguments, claims, summary) reflects only the extracted segment.
When a podcast signal’s thesis reads narrowly, that is expected — it describes the segment Gildea extracted, not the episode as a whole. Treat the decomposition as authoritative for what the signal covers; treat the source metadata as a pointer to where it came from.
What gets ingested
Gildea ingests from a curated set of 500+ sources across these formats:
| Format | Examples |
|---|
| Essays | Blog posts, newsletters, long-form analysis |
| News articles | Industry reporting, first-party announcements |
| Podcasts | Transcribed and decomposed |
| SEC filings | 10-Ks, 10-Qs, 8-Ks, proxy statements |
| Research papers | Academic and industry research |
| Social media | Expert discourse and announcements |
| Earnings transcripts | Quarterly earnings calls |
| Press releases | Product launches, partnerships, policy announcements |
Two types of signals
Every signal is classified into one of two content types:
Event signals
Key development signals sourced from global reporting and first-party announcements.
| Decomposition | Description |
|---|
| Summary | What happened, broken into individual sentences |
| Claims | Specific factual assertions extracted from the source |
Expert analysis signals
Expert analysis signals from leading researchers, operators, investors, and analysts who explain what developments mean.
| Decomposition | Description |
|---|
| Thesis | The author’s central thesis, broken into individual sentences |
| Arguments | Supporting reasoning lines, each with its own sentences |
| Claims | Specific verifiable facts extracted from the article, served as a flat list at the article level |
Decomposition structure
Expert analysis signals
Event signals
Text units
All decomposed pieces are text units. This is the fundamental atom of Gildea’s data model.
| Unit type | Found in | Scoring mode |
|---|
thesis_sentence | Expert analysis signals | regression |
argument_sentence | Expert analysis signals | regression |
summary_sentence | Event signals | regression |
analysis_claim | Expert analysis signals | nli |
event_claim | Event signals | nli |
- A verification verdict (only
pass is served)
- A scoring mode (
nli for claims, regression for sentences)
- A primary score (method-specific, do not compare across modes)
- An embedding (1024-dim Cohere
embed-english-v3.0 vector). Not in default responses — opt in with include=embeddings on signal detail. See Embeddings.
Claims and sentences use different scoring methods. Do not compare primary scores between nli and regression modes.
Signal card vs. signal detail
The list endpoint returns lightweight signal cards with metadata and a decomposition_text preview (thesis or summary text).
The detail endpoint returns the full decomposition tree with all verified text units. Use include=evidence for source evidence snippets.
See Progressive Disclosure for the full pattern.