Claude

Tracks the closer role and bullpen pecking order across all 30 MLB teams — who owns the ninth-inning job today, who is next in line if the current closer falters (the handcuff), and who carries DFA or demotion risk. Emits a per-reliever `save_role_certainty` signal (0-100) and flags speculation-worthy handcuffs for waiver bids. Use when the user mentions "closer", "save role", "handcuff", "ninth inning", "bullpen depth", lost save, blown save, committee, or when the waiver analyst needs to decide whether to spend FAAB on a backup reliever. This league uses SV as one of its five pitcher categories, but SV is also the most volatile and most punt-worthy cat, so tracking should always be paired with a punt-the-cat fallback recommendation.

Validates and persists signal files to the yahoo-mlb signals directory. Every MLB skill calls this skill before writing a signal. Enforces the signal-framework.md schema -- required YAML frontmatter fields (type, date, emitted_by, confidence, source_urls), range-checks numeric signals (0-100 unipolar, -100 to +100 bipolar), verifies variant_synthesis metadata, and enforces file naming. On validation failure, does not persist and routes a failure entry to mlb-decision-logger. Use when an agent or skill needs to emit a signal, validate a signal file, write to signals/YYYY-MM-DD-<type>.md, or check signal frontmatter. Triggers — "emit signal", "validate signal", "write signal file", "signal frontmatter".

Rewrites expository prose for a reader meeting the subject for the first time, not an insider who already shares the writer's knowledge. Enforces introduce-before-use (every entity, term and acronym defined on first mention, in an order a newcomer can follow), grounds or cuts every metaphor in the same breath it appears, converts compressed or nominalized logic into a plain if-then, and kills the two loudest machine tells, the em dash and the terse two-beat antithesis ("X. Not Y."). Carries an honesty pass so the prose never out-claims its evidence, marking inference as inference and preserving every number exactly. Use when a draft assumes knowledge the reader lacks, reads as insider or "clever", name-drops terms before defining them, leans on unexplained metaphor, or sounds AI-written; or when the user mentions first-time reader, plain language, define before use, sounds like AI, em dash, too clever, or jargon. Complements ladder-of-abstraction, slop-detector and narrative-fidelity-audit.

Systematically evaluates claims by triangulating sources, rating evidence quality (primary/secondary/tertiary), assessing source credibility, and reaching confidence-rated conclusions to prevent confirmation bias and reliance on unreliable sources. Use when verifying claims before decisions, fact-checking statements, conducting due diligence, evaluating conflicting evidence, or when user mentions "fact-check", "verify this", "is this true", "evaluate sources", "conflicting evidence", or "due diligence".

Methodology for building a properly structured Claude Code skill (SKILL.md + optional resources + evaluation rubric) from extracted components. Assesses complexity (Level 1 simple - SKILL.md only; Level 2 moderate - SKILL.md plus 1-3 resources; Level 3 complex - SKILL.md plus 4-8 resources), plans the resource grouping, drafts SKILL.md following Anthropic's authoring best practices (concise frontmatter with what + when triggers, body under 500 lines, progressive disclosure to resource files one level deep, workflow checklists), and constructs an evaluation rubric. Use when extracted components and a synthesis verdict are in hand and the next step is to materialize them as a skill. Trigger keywords - construct skill, build SKILL.md, skill construction, skill scaffolding, generate skill files.

Walks a finished document line by line, finds every word and phrase standing in for a concept, and breaks each one down until a reader with no background can follow it. Targets the assumed-knowledge phrase, the kind that sounds ordinary while hiding everything that matters ("moat", "flywheel", "product-market fit", "at scale"), not merely the rare word. Explains each term where it is used, then re-reads its own explanation and unpacks any new term that explanation introduced, repeating until every word bottoms out in ordinary language. Produces a term ledger recording what was flagged, how it was handled, where each explanation came from, and how many passes it took. Use after drafting any report, explainer or analysis for a non-expert audience, or when asked to explain something from first principles, assume the reader knows nothing, cross-question the jargon, define every term, or make a document readable by a beginner.

Writes the canonical substacker shared-context/section-map.md after writer confirmation of review-artifact proposals. Atomic write with backup snapshot. Validates schema before writing. Use as the final step of a Curator run, only after writer has accepted/modified proposals. Trigger keywords — update section map, write section map, commit sections, apply changes.

Attacks a finished valuation the way a skeptical reviewer would — hunts bias and verdict-first reasoning, checks that each risk is charged exactly once, tests whether growth is paid for, interrogates terminal value, reverse-engineers what the market price already assumes, and applies the seven deadly sins of acquisition analysis. Use when reviewing a completed valuation, red-teaming or challenging assumptions before acting on a number, auditing a DCF someone else built, or checking double counting, control premiums, synergy claims, comparable sets and route constraints.

Appends FIFA World Cup Fantasy decisions to tracker/decisions-log.md — logging not just what the manager chose but the full OPTION SET they chose from, which option the board recommended, whether the manager overrode it, and the manager's stated reasoning. Also archives the full generation (population → offspring → board → pick) to generations/, and updates the archetype scoreboard (which genotypes' blocks were chosen/won) and the manager's revealed preferences. Append-only; never overwrites. The override rows are the system's most valuable learning signal. Use after the manager picks an option off a board.

Builds the structural support for the in-round levers in FIFA World Cup Fantasy — scores how the squad's players are distributed across the round's match days (a good build spreads them, NOT stacked on one day), orders the 4 bench players by kickoff time (latest highest, so they stay promotable after early results land), and generates conditional manual-substitution triggers ("if early starter scores ≤ T and a later bench player hasn't kicked off, promote them"). Implements BB6 (matchday spread) and MB3/MB4 (bench order + sub triggers), and the manual-substitution rule in league-config.md. Returns a kickoff_spread assessment with over-stack flags, a ranked bench_order (1–4), and a sub_triggers list. Use when laying out a matchday plan or vetting a candidate squad's timing layout — called by wc-matchday-tactician and the strategists, especially A6 Matchday Mechanic.

Applies the Heath brothers' SUCCESs model (Simple, Unexpected, Concrete, Credible, Emotional, Stories) to make messages memorable and persuasive, with systematic analysis, targeted improvements, and scoring (0-18 stickiness scorecard). Use when making messages more memorable or compelling, preparing presentations, crafting pitches or campaigns, or when user mentions stickiness, making ideas stick, persuasion, SUCCESs framework, or Heath brothers.

Verifies every paper or named research result cited in a substacker draft uses the inline "Author(s), Institution, Year" form per style-guide, not a bare hyperlink or title-alone reference. Flags bare-hyperlink citations and missing-institution attributions. Use whenever the draft references external research. Trigger keywords — citation, paper citation, bare hyperlink, authors, institution, reference format.

Fetches PubMed articles posted in a given date window matching a keyword set, returning normalized records (PMID, title, authors, abstract, journal, date, DOI, URL). Uses NCBI E-utilities (`esearch` + `esummary` + `efetch`) via WebFetch for portability, but prefers a PubMed MCP server's `search_articles` + `get_article_metadata` tools when one is connected (faster, paginated, parsed). Builds date-bounded queries with optional MeSH expansion. Domain-neutral - usable for any PubMed scan, not just one project. Use when user mentions PubMed, NCBI, last-N-days PubMed, weekly PubMed scan, MeSH search, or when a literature-scan agent needs structured PubMed records.

For each simplified-boundary claim, drafts a one-paragraph suggestion for how to acknowledge the boundary inside the post — usually a single sentence or "but" clause — so the break becomes a teaching moment rather than hidden fragility. Runs for exactly the claims classified as simplified-boundary. Skip for all other classifications. Use after cross-reference-claim. Trigger keywords — boundary break, fold break into post, feature not flaw, simplified-boundary.

Drafts a proposed diff to substacker shared-context/goals.md showing which lines to add, remove, or change based on the quarter's review. Never writes to goals.md directly — writer applies manually. Used once per Growth Strategist review. Trigger keywords — goal reset, goals diff, update goals, goals proposal, rework goals.

Evaluates GraphRAG systems across knowledge graph completeness, retrieval relevance, answer correctness, reasoning depth, and hallucination prevention. Provides structured evaluation frameworks, metric selection guidance, and testing protocols. Use when evaluating GraphRAG quality, benchmarking multi-step reasoning, measuring hallucination reduction, or when user mentions evaluate GraphRAG, quality metrics, answer correctness, test my GraphRAG, or measure RAG performance.

Loads and validates a weekly Substack CSV stats export for the substacker Growth Analyst. Reconciles header against expected-columns schema, parses post rows + subscriber aggregates, moves file into corpus/stats/ on success, emits schema-warning stub on header drift. Never reads subscriber emails row-by-row — aggregates only. FALLBACK path when fetch-substack-stats (Chrome automation) is unavailable. Use when a CSV appears in inbox/substack-stats/, when Chrome is not logged in, or when the writer prefers manual export. Trigger keywords — CSV, Substack export, stats export, schema validation, subscriber data, manual export, CSV fallback.

Appends structured decision entries to the yahoo-mlb decision log (tracker/decisions-log.md) on behalf of any agent in the MLB team. Validates entries against the authoritative schema, serializes concurrent writes from parallel agents, and runs the Monday calibration pass to fill in outcomes and update the variant scoreboard. Use when any MLB agent needs to record a decision, when the coach requests "log decision", "append to decision log", "record variant outcome", or runs the "calibration pass".

Deep-dive analysis of a single MLB player (hitter or pitcher) for the Yahoo Fantasy Baseball 2K25 league. Web-searches FanGraphs (ATC projections), Baseball Savant (xwOBA/xBA/xERA), MLB.com (lineups, probables), RotoWire (weather, injuries), and RotoBaller (closer depth) to produce the full set of structured player signals defined in the signal framework. Emits form_score, matchup_score, opportunity_score, daily_quality, regression_index, obp_contribution, sb_opportunity, role_certainty for hitters and qs_probability, k_ceiling, era_whip_risk, streamability_score, two_start_bonus, save_role_certainty for pitchers. Use when you need to analyze player, compute daily_quality, compute regression index, produce player signals, run a hitter analysis, run a pitcher analysis, or prep start/sit inputs for the lineup optimizer.

Computes the full impact of a proposed MLB fantasy trade across all 10 H2H categories (R/HR/RBI/SB/OBP, K/ERA/WHIP/QS/SV), rest-of-season dollar value, positional flexibility, slot-value optionality, adverse-selection prior, and weeks 21-23 playoff impact. Produces a signed verdict (accept / counter / reject) with rationale and a specific counter if applicable. Use when user mentions "trade evaluation", "trade value", "should I accept", "trade delta", "counter offer", or pastes in a trade proposal from Yahoo. Defaults to COUNTER in the middle band — pure REJECT is reserved for clearly predatory offers.

For a given fantasy week (Monday-Sunday), identifies every starting pitcher scheduled to start twice, validates both probable starts, grades each matchup against the league's Quality Starts (QS) scoring rules, and ranks the list by streamability_score. Flags bullpen-game and opener risks that nearly never produce QS. Use when user mentions "two-start pitchers", "weekly streaming", "Monday-Sunday pitcher plan", "double start", "2-start SP", or preparing the weekly streaming plan on Sunday nights.

Scans the substacker published corpus for clusters of posts that could become a product — a course, a book, a cohort, or a consulting offer. Produces at most 2 candidates with evidence + audience signal, or an honest "not yet" verdict if nothing qualifies. Typically fires once the writer has 30+ posts. Trigger keywords — product hiding, course from essays, book from essays, corpus to product, product scan.

Verifies the section-break style in a substacker draft matches the post register — asterisks (* * *) for essayistic posts under 2500 words, H2 for methodology / how-to / technical posts. Flags mixed registers (H2 in a reflective essay, asterisks in a structured how-to). Per the style-guide rhythm rule. Use every draft. Trigger keywords — section break, asterisk, H2, headers, register, essayistic vs methodology.

Stress-tests a proposed analogy by finding the edge where the mapping breaks, then frames that break as a teaching opportunity the writer can fold into the post. Every analogy has a boundary; the writer's style treats that boundary as a feature. Use after generate-analogy-set and map-analogy-to-concept, for each framing. Trigger keywords — where does it break, stress-test, boundary, edge case, fold the break, analogy limits.

Domain-neutral methodology for the second level of Adler-style reading - understanding what a document is about *as a whole* and how its parts relate. Classifies content (practical vs theoretical; sequential / categorical / structured / hybrid), states unity in one sentence, enumerates major parts and their organization, and defines the problems the document tries to solve. Reusable across any extraction workflow - skill creation from a methodology document, Pass-2 content grasp on an academic paper, structural review of a long-form document. Use when an agent has done inspectional reading and now needs to map structure before deeper component extraction. Trigger keywords - structural analysis, document structure, state unity, enumerate parts, Adler Level 2, content classification, define problems.

Maintains substacker shared-context/topic-ledger.md as an append-and-update index of all topics in the corpus. Each topic row tracks seed/draft/published counts, last-touched date, top-3 seed ids by density, and a hot/warm/cold temperature indicator. Use after any seed is created, promoted to draft, published, or killed. Trigger keywords — ledger, topic index, update index, topic ledger, hot/cold topics.

Assembles a finished company analysis into the deliverable a decision-maker reads — the executive summary, the value bridge, the assumptions that matter, the sensitivity range, and the recommendation with its catalysts and risks. Use when writing up a valuation, DCF, acquisition, IPO, restructuring or corporate finance analysis, when drafting REPORT.md or the verdict, or when deciding what belongs in the report.

Ranks a proposed set of framings against the writer's voice profile, especially the analogy-direction priority — biology > organizational > sports, with physics/military as voice violations. Produces a tier rating per framing and flags any framing that would break voice. Use in the Intuition Builder pipeline after generating framings, to order them by fit with the writer's register. Trigger keywords — voice fit, analogy direction, biology to AI, organizational to multi-agent, sports to calibration.

Reads and updates the FIFA World Cup Fantasy tournament state machine (footballfantasy/context/tournament-state.md) — the temporal backbone tracking phase (pre-tournament → group MD1-3 → R32 → R16 → QF → SF → final), budget ($100m group / $105m knockouts), nation cap (3 group, loosening in knockouts), chips remaining, surviving nations, each owned player's elimination-risk horizon, and deadlines. Validates state on load (count/feasibility checks), applies phase transitions, and appends to the append-only state log (never silent overwrite). Use to load state at the start of a run and to commit state changes after the manager makes a move.

Plans between-round FIFA World Cup Fantasy transfers — budgets the round's free transfer(s), forces out players whose nation has been eliminated, chases fixture-swing drops, upgrades on value, and decides when a rebuild is large enough to fire the Wildcard instead of spending free transfers one at a time. Ranks candidate in/out pairs by EV gain over each player's remaining survival horizon (delta xEV weighted by progression_carry) MINUS transfer cost (a free transfer is cheap, a points hit is real, churning the squad for marginal swings is a critic flag), and tags forced/fixture/upgrade priority. Emits a `transfer-plan` signal. Use when called by wc-squad-architect (whose transfer work this skill is the engine for) and by the strategists in the populate stage when their candidate is transfer-adjacent rather than a full rebuild.

Renders the Trend Scout ranked keep-and-drop lists into ops/trend-scout/YYYY-WW-digest.md using the agent voice profile and including an appendix of all sources surveyed. Use once per weekly run as the terminal skill. Trigger keywords — weekly digest, write digest, Trend Scout digest, Saturday morning digest.

Composes the substacker final ops/growth-analyst/YYYY-WW-report.md from ingest + baseline + attribute + per-section + public-page outputs. Enforces 400-800 word budget, YAML frontmatter schema, seven-section body structure. Truncates weakest sections first when over budget. Injects data-caveats from any degraded-mode flags upstream. Use as the final compose step of the weekly pipeline. Trigger keywords — weekly report, compose report, growth report, Monday report.

For every analogy in a substacker draft, verifies it carries mechanical weight — the analogy does real work explaining the mechanism, not merely decorates it. Cross-references analogy-catalog.md for novelty (is this analogy reused from a prior post?) and domain fit (biology > organizational > sports preferred; physics/military disfavored). Use whenever an analogy appears in the draft. Trigger keywords — analogy weight, decorative, mechanical weight, reused analogy, catalog check, metaphor check.

Projects 30-90 days of daily cash flow per cash account by combining current balance, scheduled recurring inflows and outflows from recurring.json, and a 6-month rolling average of discretionary spend, then flags days where projected balance falls below a configurable safety floor. Use for forward planning, detecting upcoming overdrafts, sizing pre-funding for sinking funds, or when user mentions cash flow projection, runway, balance forecast, or upcoming bills coverage.

Domain-neutral methodology for the third level of Adler-style reading - extracting structured components (terms, propositions, arguments, solutions) from a document section by section. Selects a reading strategy appropriate to document size and structure (section-based for documents under 50 pages with clear sections; windowing for long documents without breaks; targeted for hybrid content where only specific sections matter). Writes per-section extraction notes that downstream synthesis can consume. Reusable across any extraction workflow - skill creation from a methodology document, Pass-2 content grasp on a paper's full text, evidence-mining from a long-form report. Use when an agent has done structural analysis and now needs to extract the actual atomic content. Trigger keywords - component extraction, section-by-section extraction, extract terms, extract propositions, extract arguments, Adler Level 3, interpretive reading.

Runs an intrinsic (discounted cash flow) valuation from value drivers — builds the year-by-year forecast with fading growth, margins and reinvestment, closes it with a terminal value, adjusts for failure risk, walks the bridge from operating assets to equity value per share, and runs sensitivity grids. Also reverse-engineers the market price into the growth or margin it already assumes. Use when valuing a company intrinsically, building a DCF or FCFF/FCFE model, computing terminal value or value per share, running valuation sensitivity, or asking what the market is pricing in.

Guides structured identification of potential harms, benefits, and differential impacts across stakeholder groups for decisions affecting people. Covers stakeholder mapping, fairness evaluation, risk mitigation design, and monitoring. Use when decisions could affect groups differently, need to anticipate harms/benefits, assess fairness and safety, identify vulnerable populations, or when user mentions ethical review, impact assessment, differential harm, safety analysis, bias audit, or responsible AI/tech.

Extracts the 5-7 point argument backbone of a published substacker essay into a structured _spine.json working artifact that downstream platform-rewrite skills consume. Pulls verbatim sentences where possible (not paraphrases). Tags each point with evidence anchor (paper, anecdote, formula, analogy), essay section, and translatability score. Use at the start of a Distribution Translator run. Trigger keywords — spine, backbone, extract claims, thread spine, argument skeleton.

Pulls substacker's weekly Substack stats directly from the dashboard via Claude-in-Chrome browser automation. Navigates to substack.com/stats, parses the posts table and subscribers table, and produces the same typed WeekExport object that ingest-substack-csv produces — but without requiring a manual CSV export. The writer keeps Chrome signed in to Substack; this skill opens the dashboard in a new tab, reads the rendered stats, closes the tab. Primary data path for the Growth Analyst; ingest-substack-csv is the fallback when browser automation is unavailable. Trigger keywords — fetch stats, Substack dashboard, auto stats, Chrome stats, dashboard scrape, live stats, no CSV.

Generates exactly 5 distinct intuitive framings for a given technical topic — one everyday analogy, one physical metaphor, one contrarian take, one historical angle, one counterfactual. Each framing is a short scaffold (not prose), paired with its archetype and a one-line framing statement. Use when the writer invokes the Intuition Builder agent, as the core generation step before mapping, stress-testing, novelty checking, and voice fitness. Trigger keywords — generate framings, analogies for, give me 5, intuitive angles, framing set.

Domain-neutral methodology for the first level of Adler-style reading - systematic skimming to determine what kind of document this is, what it's about as a whole, and whether deeper engagement is worth the time investment. Read title, metadata, table of contents or section headings, abstract or introduction, conclusion, and end-material at a glance. Classify document type (methodology, framework, tool, theoretical, reference, or hybrid). Decide whether to escalate to deeper reading. Reusable across any artifact-from-document workflow - paper extraction, skill creation from a methodology document, literature triage, reading-list pruning. Use when an agent needs to convert "I have a document, what is it" into structured downstream-actionable input. Trigger keywords - inspectional reading, systematic skimming, document classification, Adler reading, skim before reading, skill-worthiness check, paper triage, reading triage.

Decomposes any ML construct (attention, layer norm, softmax, convolution, dropout, contrastive loss, diffusion step, gradient descent update, cross-entropy) into the small set of linear-algebra primitives it's built from, then explains why those primitives produce the observed behavior. Includes an ablation thought experiment that asks "what would break if you removed this piece?". Use when the user asks "why does X work?", "explain attention/conv/norm intuitively", "what's actually happening in this layer", or when reading an architecture paper and a particular block doesn't make sense yet.

Analyzes a single MLB game from a fantasy perspective given home team, away team, and date. Emits structured matchup signals -- opp_sp_quality, park_hitter_factor, park_pitcher_factor, weather_risk, bullpen_state -- and a short narrative of platoon implications (handedness matchup for hitters). Use when preparing daily start/sit calls, evaluating a streaming pitcher's environment, sizing weather risk, or when user mentions matchup analysis, park factor, opposing pitcher, weather risk, or platoon.

Builds the business story behind a valuation and connects it to the drivers that carry value — market size, growth, margin, reinvestment efficiency and risk. Tests a narrative as possible, plausible and probable, and keeps the feedback loop open. Use when starting a valuation, writing the story behind a company, setting growth or margin assumptions, sanity-checking a story against its numbers, choosing a target margin or sales-to-capital ratio, or when a valuation needs a defensible reason for its inputs.

Writes quantities into prose so they land without overstating what the source supports. Applies a per-paragraph number-landing checklist (three numbers per paragraph, one per sentence, direction before magnitude, a comparison on the same screen), a unit-and-denominator lock that catches denominator drift, quantitative fidelity refusals (precision ceiling, rounding drift, percent vs percentage point, model output vs measurement), an uncertainty routing table, a declared likelihood-and-confidence ladder, and a chart-beat contract for exhibits that carry a claim. Use when drafting or editing prose that carries figures, statistics, benchmarks, forecasts or estimates, when a passage reads as a stat wall, when deciding how much precision or hedging a number can bear, or when the user mentions numbers in prose, rounding, denominator, per capita, percentage points, error bars, confidence, uncertainty, chart title, caption, or stat wall.

Parses a single financial statement PDF (checking, savings, credit card, brokerage, 401k, HSA, mortgage, tax form) and emits a normalized JSON record with institution, account mask, statement period, opening/closing balances, line-item transactions or holdings, and a confidence score. Use when extracting structured data from a bank, brokerage, retirement, or HSA PDF statement, when ingesting a drop of household finance documents, or when user mentions parsing a statement, extracting transactions from a PDF, or normalizing statement data.

Breaks down weekly and trailing-4-week substacker performance per Substack section, keyed on section tags in corpus/published/ and section-map.md. Reports opens, clicks, subs-attributable-to-section per section with ≥3 posts. Skips if section-map has <2 sections. Feeds Curator with pruning candidates (sections with 4-week median z ≤ -1.0). Use when section-map has ≥2 live sections. Trigger keywords — per-section, section performance, section metrics, section pruning, differential engagement.

Runs a voice-fidelity audit on each of the four substacker platform outputs (Substack Note, X thread, LinkedIn post, cross-post blurb) before reporting Distribution Translator completion. Checks for voice-don'ts (banned vocabulary, emoji, generic openers, marketing math without source), voice-do compliance (paper attribution preserved, hedges preserved, em-dash reframes present), platform-specific tonal shifts. Emits voice-check.md with pass/fail per artifact. Trigger keywords — platform voice check, voice-check, gate, distribution voice, slop-leak check.

Turns a raw dictation transcript into a clean, editable draft while preserving the writer's voice and form exactly. MECHANICAL ONLY — removes fillers and false starts, restores sentence-boundary punctuation and paragraph breaks, and FLAGS likely transcription errors as questions rather than silently fixing them. It does NOT restructure, restyle, reword, or improve. Produces the cleaned draft plus a cleanup log of every change and every flag. This is the strict intake bridge from "speak out my article" to an editable draft, run before the advisory-edit passes. Use when cleaning a dictation/voice transcript, prepping a spoken draft for editing, or doing transcript intake. Trigger keywords — clean up transcript, dictation cleanup, spoken draft, transcript intake, remove fillers, flag transcription errors.

Verifies that an extracted statement is internally consistent by checking that opening_balance + sum(transactions) = closing_balance within a small tolerance, and produces a reconciliation report flagging missing rows, double-counted rows, sign errors, and rounding diffs. Use as the gate between PDF extraction and committing transactions to the data store, when reconciling a statement that does not balance, or when user mentions reconcile, balance check, or statement does not tie out.

Renders a market, protocol, institution, codebase, supply chain, disease, or ecosystem as a legible protagonist without faking agency. Builds the POSIWID character sheet (revealed objective function vs. stated purpose), the agency ledger of verb repairs, the gateway-proxy admission test, the traveling-object device, the instrumented threshold scene, and climax by constraint violation. Use when the subject of a piece is a system rather than a person, when a draft says "the market decided" or "the industry learned", when analysis reads as a data dump with no protagonist, or when user mentions systemic narrative, system as character, POSIWID, anthropomorphism, agency laundering, follow the flow, or institution as protagonist.

Computes clean-sheet probability and defensive-returns EV for a team in a specific fixture from team defensive strength (xGA) against opponent attacking strength (xG), then models the STACK CORRELATION that turns a Clean-Sheet Spine (BB2) — GK + k same-team defenders — into a single correlated bet rather than k independent draws. Returns p_cs(team,fixture), a stack_corr_bonus capturing the amplified joint ceiling and shared downside (sign interacts with the rank objective θ — amplify when chasing, discount when protecting), and expected_ga with its conceding-points downside. Use when sizing a defensive stack, valuing a GK/DEF pick, scoring a candidate's BB2, or weighing the Clean Sheet Shield chip — called by wc-squad-architect, wc-matchday-tactician, wc-fitness-eval, and the strategists.

Scans transactions for fraud and anomaly signals — duplicate charges within 48 hours, transactions more than 3 standard deviations above a merchant's historical average, first-ever transaction with a new merchant above a high-dollar threshold, and unusual geography or time. Produces severity-tagged alerts with the transaction id, evidence, and a recommended action (call bank, freeze card, dispute, monitor). Use for vigilance scans on every drop, after any large unexplained outflow, or when user mentions fraud check, suspicious charge, anomaly detection, or duplicate charge.

For each substacker post that materially over- or under-performs the rolling baseline (|z| ≥ 1.0), produces a plain-English attribution paragraph with calibrated confidence (high / medium / low / unexplained). Considers subject-line effect, topic zeitgeist, external share, day-of-week, length effect, and audience-notes signals. Labels unexplained outliers explicitly rather than fabricating a story. Use after compute-baseline when outlier posts exist. Trigger keywords — attribution, why did this post work, outlier explanation, performance analysis.

Cross-references each proposed analogy in a 5-framing set against substacker shared-context/analogy-catalog.md to flag reuse. Classifies each analogy as new, reused-from-catalog (and which entry), or adjacent-to-catalog (close to an existing entry but not identical). Prevents the writer from recycling "imagine a library" for the twentieth time. Use after generate-analogy-set and before presenting framings to the writer. Trigger keywords — novelty, catalog, analogy reuse, already used, imagine a library.

Cluster a conference's event records into a small set of coarse themes with finer sub-clusters, an explicit outlier bucket, and soft (multi-membership) affinities — using the hybrid embed-then-label pipeline (embed abstracts, reduce, density-cluster, then LLM-label the clusters) when embedding libraries are available, and an LLM-reasoned hierarchical fallback when they are not. Embeddings do the grouping; the LLM only names the groups. Conference-agnostic. Use when turning structured event records into a navigable theme map for preference elicitation and scheduling, when you need 6-8 reasonable themes rather than 20 muddy ones, or when overlapping talks must belong to more than one theme. Trigger keywords - theme clustering, cluster talks, embed then label, soft membership, outlier talks, conference themes, topic map.

Finds and cites a primary source for each substacker draft claim — original arXiv paper, official documentation, RFC, canonical textbook. Source hierarchy enforced — primary > secondary > tertiary > not-a-source. Records URL, title, passage/result that settles the claim. Use after classify-claim; runs once per claim unless classification is simplified-correct with high confidence on standard undergrad material. Trigger keywords — cross-reference, primary source, citation, arXiv, RFC, paper lookup, source hierarchy.

Checks a candidate seed against the existing substacker corpus (seeds, drafts, published) for exact duplicates (sha256 fingerprint) and near-duplicates (title Jaccard, first-200-word Jaccard, shared topic cluster). Exact match exits as SKIPPED. Near-match links via related_seeds rather than creating a duplicate. Use after topic tagging and density scoring, before writing the seed. Trigger keywords — dedupe, duplicate, already thought, near-match, related seed, fingerprint.

Generates structured experimental designs (factorial, response surface, Taguchi) to systematically discover how multiple factors affect outcomes while minimizing experimental runs. Use when optimizing multi-factor systems with limited experimental budget, screening many variables to find the vital few, discovering interactions between parameters, mapping response surfaces for peak performance, validating robustness to noise factors, or when users mention factorial designs, A/B/n testing, parameter tuning, or process optimization.

Provides structured formats and techniques for running productive group sessions, from standups to multi-day workshops. Covers format selection, agenda design, participation management, decision methods, and handling difficult dynamics. Use when running meetings, workshops, brainstorms, design sprints, retrospectives, or team decision-making sessions, or when user mentions facilitation, workshop design, meeting patterns, session planning, or effective collaboration.