Skill Claude CodeCodex
Checks new prompt versions for unsafe patterns, format drift, registry consistency.
Multi-agent (LangGraph + Claude) app that turns an earnings call into a source-attributed analyst brief: ingest - FinBERT tone + KPI-vs-consensus - SurpriseSignal - grounding check - delivery.
Skill Claude CodeCodex
Checks new prompt versions for unsafe patterns, format drift, registry consistency.
Skill Claude CodeCodex
Generates a concise Markdown analyst brief from the synthesized signal and supporting evidence. Use whenever a human-readable earnings brief must be produced for delivery to Slack/Notion/email.
Skill Claude CodeCodex
Delivers the finalized analyst brief to Slack, Notion, Google Sheets, and Gmail via authenticated Composio tools. Use whenever a completed, grounded brief must be distributed and logged.
Skill Claude CodeCodex
Retrieves analyst consensus EPS and revenue estimates from yfinance for surprise calculations. Use whenever actual results must be compared against the street's expectations.
Skill Claude CodeCodex
Computes EPS and revenue surprise percentages from actuals versus consensus. Use whenever a beat/miss magnitude is needed as a quantitative signal input.
Skill Claude CodeCodex
Scores financial text sentiment (positive/negative/neutral) using the FinBERT model, aggregated to a single tone score per segment. Use whenever earnings text, prepared remarks, or Q&A needs calibrated finance-domain sentiment rather than generic LLM sentiment.
Skill Claude CodeCodex
Verifies that every claim in the analyst brief is supported by the source transcript and structured KPIs using an LLM-as-judge, returning a faithfulness score. Use whenever a generated brief must be checked for hallucination before delivery.
Skill Claude CodeCodex
Detects whether management raised, lowered, or maintained forward guidance by comparing stated guidance to the prior consensus for the next period. Use whenever forward-looking guidance direction is a required signal.
Skill Claude CodeCodex
Extracts linguistic hedging and certainty features (uncertainty words, modal verbs, certainty markers) from earnings text. Use whenever management's confidence level or evasiveness must be quantified alongside FinBERT sentiment.
Skill Claude CodeCodex
Extracts structured financial KPIs (revenue, EPS, segment revenue, guidance) from earnings text using few-shot Claude structured output. Use whenever numeric KPIs and forward guidance must be pulled from a transcript or press release into a typed schema.
Skill Claude CodeCodex
Retrieves a company's most recent 8-K current report and its Exhibit 99.1 earnings press release text from SEC EDGAR via edgartools. Use whenever the task needs an official earnings press release, 8-K event items, or a transcript fallback when earningscall has no coverage for the ticker.
Skill Claude CodeCodex
Computes a sentiment surprise by comparing the current quarter's tone against the trailing eight-quarter mean stored in long-term memory. Use whenever the analysis needs tone change relative to history, not just an absolute tone score.
Skill Claude CodeCodex
Combines sentiment surprise, hedging, EPS/revenue surprise, and guidance into a single transparent weighted score with direction and confidence. Use whenever the per-agent features must be fused into one actionable signal.
Skill Claude CodeCodex
Ingests an earnings-call transcript and segments it into prepared remarks vs analyst Q&A. Use whenever sentiment, tone, or KPI analysis needs the call split into the two sections, or when a raw transcript must be normalized before downstream agents run.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: