+ 
A public, browsable dashboard that shows exactly what changed in a company's SEC risk-factor disclosures between consecutive filings — a GitHub pull-request diff, applied to prose.
@@ -32,7 +33,15 @@ strip gives an at-a-glance sentence count of what changed before you read a word itself, a section-jump sidebar lets you skip straight to a specific risk topic instead of scrolling a 40-paragraph document, and an analyst-metrics panel surfaces three deterministic, citation-backed disclosure-analysis figures (textual similarity, Fog readability, section length) alongside the -qualitative diff. +qualitative diff, plotted as a trend across every filing the company has, not just the two periods +currently selected. + +The pre-seeded universe (~38 well-known tickers) isn't a hard limit: searching a ticker that isn't +indexed yet offers to import it live, straight from SEC EDGAR, using the exact same extraction/diff +pipeline as the offline batch job. Beyond a single company's own history, a company's numbers are +also placed against real sector peers (a percentile rank against every other same-sector company's +own latest metrics — arithmetic over data already in the database, not a model output), and a +full-text search reaches across every ingested filing's real Item 1A text at once. This is a standalone project built to demonstrate four things: **SEC EDGAR filing ingestion and section-level text extraction** (locating a specific, inconsistently-formatted legal section @@ -51,6 +60,8 @@ handling** for the one optional AI feature in the app. - [Why this exists](#why-this-exists) - [Architecture](#architecture) - [Installation / running locally](#installation--running-locally) +- [Local AI summaries (optional)](#local-ai-summaries-optional) +- [Testing](#testing) - [Quickstart](#quickstart) - [How the diffing works](#how-the-diffing-works) - [Data handling & privacy](#data-handling--privacy) @@ -107,10 +118,29 @@ flowchart LR Notice what's *not* in the second diagram: this project's own FastAPI server. There is no passthrough endpoint or proxy — the request goes straight from your browser to Anthropic. +**3. An optional local-LLM summary layer — precomputed offline, served to everyone:** + +```mermaid +flowchart LR + D[("SQLiteFirst paragraph.
Second paragraph.
" \ + "${escapeHtml(c.text.slice(0, 160))}${c.text.length > 160 ? "…" : ""}
+ ${c.words} words +Real word counts grouped under this filing's own section headings -- not a score or a judgment call.
+${escapeHtml(peerTicker)} doesn't have enough filings to compare yet.
`; + return; + } + const [thisDiff, peerDiff] = await Promise.all([ + fetchDiff(ticker, thisLatest[0].filing_id, thisLatest[1].filing_id), + fetchDiff(peerTicker, peerLatest[0].filing_id, peerLatest[1].filing_id), + ]); + const tm = thisDiff.analyst_metrics; + const pm = peerDiff.analyst_metrics; + const rows = [ + ["Textual similarity", tm.similarity_score != null ? `${tm.similarity_score.toFixed(1)}%` : "—", pm.similarity_score != null ? `${pm.similarity_score.toFixed(1)}%` : "—"], + ["Fog readability", tm.fog_index_to?.toFixed(1) ?? "—", pm.fog_index_to?.toFixed(1) ?? "—"], + ["Item 1A length", tm.word_count_to?.toLocaleString() ?? "—", pm.word_count_to?.toLocaleString() ?? "—"], + ]; + result.innerHTML = ` +| ${escapeHtml(ticker)} | ${escapeHtml(peerTicker)} | |
|---|---|---|
| ${label} | ${a} | ${b} |
Each company's own most recent filing-over-filing comparison, fetched fresh — not the pair currently selected above.
`; + } catch (e) { + result.innerHTML = `Couldn't load ${escapeHtml(peerTicker)}: ${escapeHtml(e.message)}
`; + } + }); +} + +function renderLocalSummaryPanel(local) { + if (!local) return ""; + const date = new Date(local.generated_at); + const dateStr = Number.isNaN(date.getTime()) ? local.generated_at : date.toLocaleDateString(); + return ` +${escapeHtml(local.summary)}
+Generated offline by a local model run against this diff on ${escapeHtml(dateStr)} -- served to every visitor, not tied to any API key. Not guaranteed to be accurate; verify against the original filings.
+Generating summary…
${escapeHtml(text)}
Generated by Claude from the diff above. Not guaranteed to be accurate -- verify against the original filings linked above.
`; }) .catch((e) => { panel.innerHTML = ` -Couldn't generate a summary: ${escapeHtml(e.message)}
`; }); } diff --git a/frontend/js/home.js b/frontend/js/home.js index 6f7d016..8924763 100644 --- a/frontend/js/home.js +++ b/frontend/js/home.js @@ -1,8 +1,8 @@ -import { fetchCompanies, fetchMeta } from "./api.js"; +import { fetchCompanies, fetchCompany, fetchDiff, fetchMeta, importCompany, searchCompanies, searchFilingText } from "./api.js"; import { countUp } from "./animate.js"; function escapeHtml(s) { - return s.replace(/[&<>"']/g, (c) => ({ + return String(s).replace(/[&<>"']/g, (c) => ({ "&": "&", "<": "<", ">": ">", '"': """, "'": "'", }[c])); } @@ -11,30 +11,24 @@ function magnitudeOf(c) { return c.latest_change ? c.latest_change.added + c.latest_change.removed : -1; } -function companyCard(c, i) { +function companyCard(c) { const change = c.latest_change; - const changeHtml = change - ? `Browse how well-known public companies' disclosed risk factors (Item 1A) have - changed, filing over filing — a plain text diff, not a prediction.
-A GitHub pull-request diff, applied to the risk factors public companies disclose to + the SEC — every added, removed, and unchanged sentence, filing over filing.
+