paper_fast_search_backup

Use this skill when the user wants questions about flowcharts, diagrams, or info hidden in page headers, footers, and complex 2D structures. Trigger it for requests like 'follow the process in this diagram', 'read the chart layout', 'what comes after this step in the map', or 'check the footer for the document code'. This is for scenarios where the answer isn't just in a text block but depends on where things are placed on the page.

Use this skill when the user provides a 'story' or hypothetical scenario and needs to find the specific legal rules or statutes that apply to it, even if the scenario and the documents share very few common words. It is triggered by requests like 'find the laws that cover this situation,' 'tell me if this person broke a rule based on the handbook,' 'apply the housing code to this tenant story,' or 'help me with a bar-exam style legal lookup.' This skill maps everyday factual descriptions to formal legal authority through reasoning-driven retrieval.

Use this skill when a user wants to find a specific time window or 'clip' within a video based on an event, object, or action. Trigger it for requests like 'find the part in the video where they score', 'get the timestamps for the interview segment', 'show me the moment the cat jumps', or 'extract the clip of the engine failure'. This skill is for navigating continuous temporal visual data to locate discrete start and end markers (grounding).

Use this skill when the user wants to test if the model can be tricked into giving out dangerous or forbidden information by using creative writing or 'acting' tricks. Trigger it for requests like 'make the agent answer a harmful question by role-playing', 'try to get past the rules by using slang or expert-sounding language', 'ask for instructions on illegal acts but frame it as a story', or 'test if the model refuses bad advice even when it's pressured to be helpful'.

Use this skill when you need to find the 'most important' or 'landmark' documents in a collection where popularity and rank matter as much as keywords. Trigger it for requests like 'find the landmark rulings', 'give me the most cited research papers', 'prioritize high-court decisions over local ones', or 'rank the results by how much other documents point to them'. This skill is essential for navigating legal, academic, or professional corpora where information is organized in a hierarchy or linked through a citation network.

Use this skill when the user wants 'connect the dots' questions, exploratory 'multi-hop' patching, or tracking fragmented clues across multiple sources or organizational channels. Trigger it for requests like 'force the model to follow a shared thread across docs', 'track this ticket back to the original slack request', 'find the missing pieces of a puzzle', or 'gather all mentions by connecting several hidden dots'.

Use this skill when the user wants to solve a medical problem by looking at what happened to other similar patients. Trigger it for requests like 'find similar cases to this patient', 'how do doctors usually treat people with these same symptoms', 'give me examples of past medical experiences for this diagnosis', or 'compare this patient's history with the hospital records to find a pattern'. This skill focuses on a coarse-to-fine retrieval process that identifies analogous entities before extracting specific decision-making evidence.

Use this skill when the user wants questions with wrong snippets, misleading facts, adversarial 'White DoS' system warnings, or answers that sound plausible but go beyond the docs. Trigger it for requests like 'put a false statement in the context', 'make the docs disagree', 'see if it follows a fake command to stop answering', or 'inject an adversarial poisoned document'. This skill is about multi-source consensus, evidence-grounded correction, and resisting hallucinatory or adversarial hijacking.

Use this skill when the user wants to generate a complex technical design or code output by selecting and assembling modular components from a pre-defined library according to specific quantitative constraints (budgets). Trigger it for requests like 'build a configuration that fits this power budget', 'pick the right hardware modules to keep area small', 'assemble a design from the library using these specs', or 'generate a project top-file that pulls in the right parts'.

Use this skill when the user wants table-heavy questions, multi-constraint mathematical logic, SQL/programming tracing, or questions requiring post-processing formats. Trigger it for requests like 'make it use a table and text', 'trace the math logic of this textbook', 'ask for the one that fits several conditions', 'cross-reference the table with the article', or 'force an algorithmic calculation'. This skill targets complex operation execution after foundational text retrieval.

Use this skill when the user asks a question in one language (e.g., German, Spanish, Chinese, Arabic) but the information needed is stored in documents of a different language (e.g., English), or split across documents in multiple languages. Trigger it for requests like 'find the answer from these English reports and tell me in German', 'compare what the local news and the international news say', or 'check if the model flips to English when I ask a question in Spanish about an English text'. This skill targets the unique challenge of performing complex reasoning (comparison, aggregation) across language boundaries while maintaining the correct response language.

Use this skill when the user wants to compare, contrast, or find differences between different sections of a document or across multiple documents. Trigger it for requests like 'compare the 2023 and 2024 results', 'how do these two authors differ in their views?', 'what changed between the beginning and the end of the story?', or 'show the differences between these two companies'. This skill is for synthesis tasks where the answer requires extracting distinct facts from separate locations and performing a logical comparison.

Use this skill when the user wants questions that 'need reading everything', 'leave no file behind', 'spread the clues across many documents', 'list all entities associated with a topic', or 'test whether the model skips scattered fragments'. Trigger it for large multi-document bundles where the answer depends on broad coverage, assembling high-cardinality lists, or piecing together small discontinuous facts spread across massive reports.

Use this skill when the user asks questions about specific articles, sections, or clauses in complex documents like laws, technical manuals, or company policies. Trigger it for requests like 'what does section 5 say about X', 'find the penalty for violating article 10', 'look up the rules in chapter 3', or 'connect different parts of the law that mention each other'. This skill is specifically for identifying information that is split across a hierarchy (Titles, Chapters, Sections) or linked via internal cross-references.

Use this skill when the user provides complex questions with multiple specific rules, such as 'List items that fit A, B, and C but not D,' or 'Write a response with exactly 5 points, each starting with a vowel.' Trigger it for requests like 'make sure you don't forget the location constraint,' 'check if I've included everything I asked for,' or 'test if the model ignores the background info I provided.' This skill ensures that every semantic role (subject, action, time, location, qualifier) defined in the user's intent is strictly honored, preventing 'Intent Hallucination' where the agent omits or misinterprets query components.

Use this skill when the user wants to find the best time for an activity based on changing conditions like weather or traffic. Trigger it for requests like 'when is the earliest I can leave to avoid rain', 'find a 2-hour dry window for my trip', 'help me postpone my travel to avoid congestion', or 'is there a better time today for a walk'. This skill guides the agent to iteratively search through time records, moving a 'search window' forward or backward until a period that satisfies all the user's constraints is found.

Use this skill when the user wants to find information trapped inside data tables, spreadsheets, or databases but doesn't know which file to open. It is perfect for requests like 'find the table with school scores in Alameda', 'which database has the sports results for 2009', 'check if this fact about racing is true', or 'write a query for the city furthest from Boston'. Use this specifically when table titles are missing, uninformative, or when the answer requires pulling from multiple related databases.

Use this skill when the user wants to test if the model can be 'tricked' by biased, one-sided, or selectively framed documents. Trigger it for requests like 'test if it falls for misinformation', 'give me documents that look convincing but are wrong', 'try to lead the model to the wrong answer with biased text', or 'check if the agent can spot partial truths and selective omissions'.

Use this skill when the user provides multiple specific clues and wants to find the one person, place, or thing that fits all of them. Trigger it for requests like 'find someone who did A and also did B', 'who is the person mentioned in both these articles', 'locate an entity satisfying several conditions', or 'which item fits this set of descriptions'. This skill is specifically for identifying a unique subject through the intersection of independent factual constraints found across different parts of the corpus.

Use this skill when the user asks a follow-up or clarification question that depends on previous conversation history. Trigger it for requests that use pronouns like 'he', 'it', or 'that', or elliptical phrases like 'what about its address?', 'why was he there?', or 'any more examples of those?'. This skill is specifically for RAG agents that need to rewrite a non-standalone user turn into a standalone search query by resolving conversational co-references and context.

Use this skill when the user wants to ensure the agent doesn't exaggerate or oversimplify facts from the text. Trigger it for requests like 'don't make it sound more certain than it is', 'keep the specific conditions', 'avoid turning a maybe into a definitely', 'be careful with opinions presented in the docs', or 'don't leave out the fine print'. This skill ensures that the generated answer preserves the exact degree of certainty, specificity, and rhetorical nuance (e.g., distinguishing an opinion from a fact) found in the original retrieved passages.

Use this skill when you want to see if the agent can tell the difference between what it 'just knows' (from its own memory) and what is actually written in the provided files. Trigger it for requests like 'tell me when the agent is guessing', 'separate facts in the doc from general knowledge', 'check if the model is hallucinating outside the search results', or 'make sure the answer stays true even if it's not in the context'.

Use this skill when the user wants the agent to sound like a natural conversationalist who keeps the dialogue alive rather than providing dead-end answers. Trigger it for requests like 'make the agent more proactive', 'add follow-up questions to the results', 'simulate a back-and-forth chat', or 'encourage the user to keep asking things'. It is specifically geared towards generating synthetic training data where the assistant answers a query and then pivots to a related, inquisitive follow-up based on the retrieved knowledge.

Use this skill when the user wants to test the agent's ability to find relevant documents that do not share many words with the question but are linked by deeper logic. Trigger this for requests like 'find a case that matches this story', 'give me law that applies to this situation even if the terms are different', 'search based on the legal issue rather than keywords', or 'test if the model can spot the underlying rule in a complex scenario'. It is specifically designed for high-reasoning legal or technical retrieval where lexical overlap is extremely low.

Use this skill when the most direct piece of information is missing from the files and the agent must find a 'backup' way to the answer. Trigger it for requests like 'find a workaround in the data', 'what if the direct link is gone', 'use indirect clues to fill the gap', 'connect the dots through a middle person', or 'verify if the model can still figure out the answer even with incomplete records'.

Use this skill when the user wants to ensure the highest degree of accuracy through a 'think-twice' approach or self-critique. Trigger it for requests like 'double check your answer against the text', 'reflect on your initial guess', 'make sure you didn't miss any contradictions between the files and your internal logic', or 'correct your reasoning if the documents say something counter-intuitive'. This skill is specifically for RAG agents that follow an iterative 'Refine' loop to improve reasoning steps by critiquing a preliminary response against retrieved evidence.

Use this skill when the user wants to test if the model 'cheats' by guessing answers when the logical steps are missing. Trigger it for requests like 'don't answer if the bridge is missing', 'check if it falls for false assumptions in the question', 'make sure it doesn't guess from the answer options', or 'test if it spots that the question is too specific for the text'. This skill focuses on resisting 'shortcut reasoning' where a model provides a correct-looking answer despite failing to trace the full evidence chain.

Use this skill when the user asks questions about how to find specific features or perform multi-step tasks within a software application. Trigger it for requests like 'where is the settings menu?', 'guide me to the logout button', 'how do I view the privacy policy in this app?', or 'show me the path to buy a stock'. This skill is specifically for retrieving structured navigation sequences from specialized knowledge bases that map user intents to application UI transition graphs.

Use this skill when the user wants 'connect the dots' questions, hierarchical path-traversals, or questions about navigating a sequential process. Trigger it for natural-language requests like 'multi-hop questions', 'find a way to get to...', 'show me the steps to...', 'force the model to follow a shared thread across docs', or 'test if it can find the path even if some instructions are missing'. It targets structural traversal across documents, procedural manuals, or interface states.

Use this skill when the user wants to synthesize data testing an agent's ability to handle contradictions, misleading facts, factual errors, or semantic drift (e.g., from AI-generated summaries). Trigger it for requests like 'make the documents disagree with each other', 'test if the model spots self-contradicting files', 'see if it gets tricked by summarized text', 'check for mistakes in rewritten news', or 'see if the agent follows a lie in the text over what it knows is true'. This skill covers factuality detection, self-contradiction, and resolving information fusion or omission.

Use this skill when the user wants table-heavy questions, specific API/formatting constraints, multi-conditional sorting, trade-off optimization, or forced use of official glossaries. Trigger it for requests like 'list items that strictly meet all these criteria', 'optimize the settings for X while staying under budget Y', 'use the official dictionary provided', 'convert the retrieved date', or 'balance accuracy with cost'.

Use this skill when the user wants questions that require reading the entire context window, ensuring no file or nested branch is left behind. Trigger it for requests like 'need reading everything', 'give me a full breakdown of this specific chapter', 'summarize the statistics for all entities found', or 'tell me all the subsections' occurrences'. It forces extreme entity density and global context accounting across both multi-document spreads and deep intra-document hierarchies.

Use this skill when you need to navigate structured knowledge bases, relational triplets, or knowledge graphs to connect different facts. It is triggered by requests like 'connect the dots between these news updates', 'find the person who meets all these shared conditions', 'what are all the projects this artist is linked to', or 'who is the owner of the company that sold the most units'. This skill emphasizes traversing explicit symbolic paths (like Subject-Predicate-Object) to find intersections or multi-step connections that a single search cannot reveal.

Trigger this skill when users want to test if an agent can say 'I don't know' or 'not enough info' when the facts are missing, semi-missing, or when the question itself is a trick. Use it for requests like 'make the model deflect from unanswerable questions', 'test if it gets caught by false assumptions', 'give me questions that ask for too much detail', or 'hide the answer and see if it guesses'. It teaches the agent to resist 'shortcut reasoning' and identify when evidence is insufficient, contradictory, or lacks the required specificity.

Trigger this skill when the answer is logically fragmented across physically non-contiguous spans or when a local entity (like a function or a paragraph) requires context from a larger structure (like a repository, class, or multi-volume report). It is used for natural language requests like 'connect the fragments from this long report', 'stitch together clues from the repository', 'summarize this function using the rest of the class for context', 'check if a second search is needed for the missing repository details', or 'compare files to find similarities'. This skill ensures the agent can proactively identify and retrieve 'anchor' documents—such as skeletons, table of contents, or related code chunks—to perform high-fidelity synthesis.

Use this skill when the user wants questions where the right clue is present but surrounded by similar-looking junk, typographical errors, or misleading formatting. Trigger it for requests like 'mix in some confusing paragraphs', 'test whether it finds the wrong year', 'hide the answer among near misses', or 'test if the model can read through OCR errors like WN8A'. This skill ensures the agent can isolate the true semantic signal amidst both macro-level topical distractors and character-level document degradation.

Trigger this skill when users provide raw scenarios, layman descriptions, or technical logs and need to identify the authoritative expert rule, clinical diagnosis, legal statute, or financial investment category (e.g., EWS Pillars) that governs or classifies it. It is activated by requests like 'troubleshoot this status report', 'classify these expenditures according to the fund pillars', 'diagnose the root cause of this observation', 'what is the governing RFC standard', or 'map this situation to a formal policy'. It emphasizes surmounting the semantic gap between informal, sensory, or conversational descriptions and the highly specialized terminology used in professional blueprints and taxonomies.

Use this skill when the user asks for 'the latest update', 'before or after', timeline sequences, finding optimal time windows, or requesting the earliest/latest feasible shift for an event. Trigger it for requests like 'find a rain-free window for my 2h trip', 'when is the latest I can leave early to avoid traffic?', 'is this API still current today?', 'was the iPad or AirTag launch first?', or 'what is the overarching trend from 2022 to 2024?'. This skill is essential for chronometric normalization, ordering evidence, and performing iterative search for time-slots that satisfy specific condition-duration constraints.

Use this skill when the user wants the agent to judge whether an answer/document is securely backed by authentic evidence, verify exact pinpoint highlight spans, validate strict antecedent document grammar (e.g. tracking 'the' to 'a'), or audit sites for safety/phishing risks and machine-generated stylistic anomalies. Trigger for 'verify this highlighted claim,' 'check if this legal clause has grammatical consistency,' 'is this link a scam?', or 'identify if any of these sentences were polished by AI.' It is essential for eliminating hallucinations, detecting structural phrasing errors, and ensuring document authenticity.

Use this skill when you need a document translated while maintaining fixed terminology, consistent pronouns, and proper formatting across the whole file. Trigger it for requests like 'translate this news report into Swahili,' 'make this English health guide available in Amharic,' 'translate this tech article but keep the paragraph structure,' or 'translate this contract while making sure the law names are correct in the target language.' It is particularly effective for low-resource languages or documents with specialized scripts where the agent must avoid repeating phrases or skipping paragraphs.