Download public agents and examples free.

Use this agent when a freelance instructor adds event details and pasted news to a Google Sheet; it produces one short KakaoTalk-ready student recruitment notice in the result column.

Use this agent when a caregiver or family member pastes family message conversations about visit schedules or elder status; it produces a very polite reply draft and can send the draft by email after setup.

Use this agent when a small food shop owner wants a short daily morning news list; it produces five practical news picks in Google Sheets for today's sales decisions.

Use this agent when a salon owner wants to compare local haircut and color prices from price board photos and existing shop prices; it produces a customer-ready Google Sheets price comparison and new price recommendations.

Use this agent when a family organizer wants to turn simple daily care notes and photo descriptions into a very short message for one family member.

Use this agent when clinic staff need to turn difficult diagnosis or test wording into a gentle KakaoTalk-ready message for a 5-7 year-old child.

Use this agent when a music teacher has lesson links, PDF files, photos, or notes and needs a clear packing checklist with reasons and sources, ready for Google Sheets, email, and printing.

Use this agent when you have a church event name, date, or official website link; it produces a very short, friendly KakaoTalk-ready notice and saves it as a file, with an email sending option after setup.

Use this agent when a music teacher has photos, screenshots, or family messages and needs a student-grouped packing checklist saved as a local Markdown file plus message-ready text.

Use this agent when care service inquiries arrive from a website form; it produces polite short reply drafts, review checklists, and Google Sheets-ready rows.

Use this agent when you paste grocery price notes from trusted local sources; it produces a simple Markdown best-price list for today's shopping choices.

Use this agent when one person wants a very short morning digest; it produces three must-know general news items ready to paste into KakaoTalk.

Use this agent when a high school student pastes short school project notes; it produces a weekly Sunday evening email checklist with carried-over unfinished tasks.

Use this agent when a school counselor wants to turn local news, school updates, city or police posts, and typed photo descriptions into a short student safety digest saved to a Google Sheet and emailed to the counselor.

Use this agent when elementary school parent event details or photo descriptions need to become one short KakaoTalk-ready notice; it produces a polished one-paragraph announcement and collects results in a document.

Use this agent when a student needs a very short daily message from one Google Sheet; it produces today's homework and supplies notice and asks before sending if anything is unclear.

Use this agent when you want to collect customer questions from a website form or Instagram and receive polite reply drafts by email.

Use this agent when a family caregiver wants to compare home care and day care prices by region using news and public guidance; it produces a short table and KakaoTalk-ready message.

Use this agent when farm sellers paste customer messages; it produces polite Markdown reply drafts grouped by customer.

Use this agent when a small bakery owner wants to compare flour and butter prices from saved suppliers, supplier message photos, and trusted online pages; it produces the cheapest supplier today, a history row, and owner alerts when a price needs checking.

Use this agent when a family uploads a photo of health readings or a short note; it produces a very simple family-friendly summary, cautions, and questions to ask a doctor.

Use this agent when clinic staff need to create a clear pre-visit checklist for a guardian; it produces a local Markdown file with preparation items, homework, and questions to confirm when information is unclear.

Use this agent when a home caregiver pastes booking or cancellation notes; it produces a reviewed table file ready for Google Sheets.

Use this agent when a grandparent wants to turn medicine package photos or handwritten appointment notes into simple family reminder rows in a shared Google Sheet.

Helps the target user run polite customer reply drafts with clear inputs, setup guidance, and safe handoff notes.

Use this agent when a local festival manager uploads receipt and memo tables; it produces a clean expense list and a KakaoTalk-ready summary.

Use this agent when a caregiver needs a simple daily checklist; it reads homework and materials from Google Sheets and Notion, then creates today's Notion checklist.

Use this agent when a cafe owner has a short photo description and wants a kindergarten-level explanation saved as a Markdown file.

Use this agent when a family wants three must-know morning news items in simple language; it produces a short KakaoTalk-ready message from pasted notes, links, and major news checks.

Use this agent when a reading group host pastes reading notes and needs a short KakaoTalk-ready explanation of difficult words for middle school students and older.

Use this agent when a student uploads a class plan or school website screenshot; it produces a simple checklist with quantities, dates, and places that can be pasted into KakaoTalk.

Use this agent when a caregiver pastes appointment or cancellation notes; it produces a short table and KakaoTalk-ready message for family members.

Use this agent when a family caregiver pastes local home care price notes; it produces a clear Notion table comparing prices, companies, regions, services, and source confidence.

Use this agent when a student pastes school notes, homework lists, or test reminders; it produces a short weekly checklist and sends it to Notion, KakaoTalk, and a web page.

Use this agent when a freelance instructor sends poster text or class photo notes; it produces a KakaoTalk-ready short notice, a detailed notice, and an email delivery copy.

Use this agent when one person needs daily medicine reminders from a Google Sheet; it sends each medicine-time reminder and tells family every time.

Use this agent when apartment office staff need a daily morning selection of important public local news and notices; it produces five news items and a KakaoTalk-ready summary text.

Use this agent when apartment office staff need a short weekly reminder based on recurring inspections, cleaning work, last week's tasks, and checklist notes; it produces copy-ready text for SMS or KakaoTalk.

Use this agent when a parent writes a short description of a child or family health record photo; it produces a very short family-friendly summary and a doctor visit note, with optional Notion saving.

Use this agent when a clinic receptionist pastes visit notes or appointment details; it produces a clear family-friendly reminder and saves the final message to Notion with optional SMS or email draft text.

Use this agent when a family caregiver pastes phone messages, copied texts, task lists, emails, or spreadsheet rows; it produces a short family update, a checklist, and a Google Sheets-ready table.

Use this agent when a non-technical person wants medicine reminders from a simple Google Sheet and Telegram; it produces short medicine-time alerts for the patient and family.

Use this agent when you paste simple school or cafe notes; it produces a clear checklist split into homework, supplies, cafe tasks, questions, and a Notion-ready result.

Use this agent when a retired neighbor wants to compare grocery prices from a Google Sheet and trusted store websites; it produces the single best current deal plus a simple report.

Use this agent when a farm stand owner wants hard words from a spreadsheet explained for ages 4-6; it produces tiny child-friendly explanations, saves them, makes a printable list, and can send them to Telegram after approval.

Use this agent when you manage medicine times in Google Sheets; it sends simple reminders, records today’s result, and lets family know.

Use this agent when a pet guardian shares a photo and short note; it produces a simple checkbox checklist for walks or vet visits and sends it to the chosen family or guardian email.

Use this agent when a high school student wants a short weekly reminder for college, career, and news reflection; it produces a Sunday evening email checklist and a saved backup copy.

Use this agent when a family member uploads a photo description of an item; it identifies the item, compares nearby public prices, and sends a short conclusion plus table to Notion once a week.

Use this agent when a soccer parent needs kind, clear suggested replies to other soccer parents; it produces a short reply, a longer reply, and sends the draft to Telegram for approval.

Use this agent when a school counselor needs to turn official school messages, website links, copied text, PDFs, or documents into a simple checklist with task and source link rows.

Use this agent when you want a simple daily medicine checklist; it produces Notion reminders, taken records, and optional family notices.

Use this agent when a clinic receptionist needs a weekday morning summary of local news from photo descriptions; it produces a short Markdown digest with patient-question notes and urgent items.

Use this agent when a small bakery owner pastes a customer message; it produces one kind draft reply and sends the draft to the owner's email for review.

Use this agent when a freelance instructor uploads classroom or material photos; it produces a short preparation checklist document with optional related news links.

Use this agent when an elementary school teacher shares a photo or related class material; it produces a simple preparation checklist and a short 3-5 page PPT draft.

Use this agent when a farm stand owner pastes daily notes or receipt text; it produces a short plain-language spending and income summary that can be reviewed and sent to Telegram.

Use this agent when a school counselor needs to turn spreadsheet rows into safe medicine and appointment reminder drafts, send a counselor test reminder, and record each reminder in Notion.

Use this agent when a grandparent wants to turn a voice note, photo, or short note into a simple travel packing checklist; it checks official travel sources first and produces an email-ready, printable, and phone-friendly checklist.

Use this agent when a book club host pastes weekly reading notes; it produces a Notion checklist with this week's reading assignment and follow-up tasks.

Use this agent when a pet sitter needs a clear neighbor-friendly lost pet, found pet, care update, or emergency notice; it produces a short announcement, an email draft, a Slack-ready message, and a Google Sheet record.

Use this agent when a family needs simple summaries of blood pressure and blood sugar records; it produces a very short family-friendly health brief and asks before sending when anything is unclear.

Use this agent when a school counselor wants to turn pasted notes, local text files, and selected school or local news pages into a privacy-first booking table saved locally.

Use this agent when a guardian pastes family chat text; it finds hard words and produces a simple child-friendly explanation table that can be sent to Notion.

Use this agent when a nonprofit coordinator needs to collect receipt details from family messages and PDFs, check unclear items first, and produce a short plain-language spending explanation that can be emailed and pasted elsewhere.

Use this agent when a college freshman pastes messy move-in notes or dorm instructions; it produces a clear local Markdown packing checklist.

Use this agent when apartment staff need to turn a Google Sheet row into a short printable homework and supplies checklist, with an email draft prepared for review before sending.

Use this agent when a student needs to draft kind customer replies from official websites and save each draft with source links to a Google Sheet.

Use this agent when a parent pastes a news article or family chat sentence; it produces a one-line summary, a child-friendly explanation, and an email-ready result.

Use this agent when a family needs simple daily summaries of health numbers and habits; it produces short plain-language notes, small tables, and checklists in Google Sheets.

Use this agent when a church volunteer leader sends a photo or short description of food ingredients or event supplies; it searches local price information and produces an easy top 3 comparison table for Telegram.

Use this agent when clinic reception staff paste appointment questions; it produces one polite reply draft ready to copy.

Use this agent when apartment office staff paste receipt notes or expense memos; it produces a simple Markdown expense table saved on the computer and shown in chat.

Use this agent when a grandparent adds a website order page; it saves the receipt, explains the spending in very short plain language, remembers past spending, and sends a daily note to Notion and a text-style chat app.

Use this agent when farm sellers paste reservation and cancellation requests into Google Sheets; it produces a clean customer-by-customer reservation table, cancellation table, check-needed list, printable summary, and phone-message summary.

Use this agent when a small shop owner pastes weekly notes and needs a simple Markdown checklist for ingredient ordering and side dish prep.

Use this agent when pet shop or animal clinic staff upload photos of existing pet items; it produces a dog or cat adoption, outing, or travel checklist with official reference links and can email the result.

Use this agent when a student needs a simple daily checklist of homework and school supplies from public school posts, teacher-provided text, and relevant news; it produces a short email, optional text message, and printable page.

Use this agent when a local plumber pastes family messages about new bookings or cancellations; it produces a simple copy-ready booking table with notes.

Use this agent when academy owners need polite reply drafts for parent, student, or partner inquiries; it produces copy-ready Korean messages for KakaoTalk.

Use this agent when an academy owner needs a daily official-source education news brief; it produces a dated Markdown file with 3 concise items and clear failure notes when sources cannot be checked.

Use this agent when a student wants to compare local prices from official pages; it produces an easy table, checklist, source notes, and Google Sheets-ready results.

Use this agent when a local hair salon owner sends price photos or links; it reads haircut and dye prices, compares nearby prices, and produces a short Telegram summary.

Use this agent when a clinic receptionist needs a daily local-news digest from article and city-alert image captions; it produces a short operations-focused summary in Notion and email.

Use this agent when a non-technical user uploads a school note or academy notice image or document; it produces a simple homework and supplies checklist and sends it to Telegram.

Use this agent when a solo shop owner wants to collect customer questions in Google Sheets and produce polite reply drafts for price and booking inquiries.

Use this agent when a real estate assistant pastes family chat messages; it extracts unfinished listing checks and client follow-ups, then produces a weekly task list and Telegram reminder.

Use this agent when a homeroom teacher receives reservation or cancellation messages; it produces a date-sorted Notion table and asks before handling unclear requests.

Use this agent when you paste a news link or article text; it produces short, easy Korean explanation text ready to copy into KakaoTalk.

Use this agent when a non-technical user wants a daily 6 AM news brief from pasted family chat interests; it produces up to 5 short news rows in Google Sheets.

Use this agent when a high school student wants 3 important morning news items; it produces a safe, sourced Notion news brief with short photo-style explanations.

Use this agent when a music teacher wants to compare lesson prices from teacher-approved local sources using photos or typed lesson names; it produces a short recommendation with source links and saved prices.

Use this agent when a reading club host pastes family chat messages; it produces a person-by-person table of reservation and cancellation requests, checks duplicates, and sends the result to Google Sheets and an in-app table.

Use this agent when a college freshman wants to save receipt messages, track semester spending, and get very short plain-language spending summaries.

Use this agent when a soccer parent wants to record team snack, drink, tournament, or game fee receipts from photo descriptions and produce a private local Markdown spending summary in very simple English.

Use this agent when a real estate office assistant has an official event page or event name and needs a short KakaoTalk-ready notice with a title, brief body, and contact line.

Use this agent when apartment office staff upload receipt files and notes; it produces a reviewed expense table with source links, Notion records, and an Excel backup.

Use this agent when you want simple morning and lunch medicine reminders; it produces a daily medicine schedule, short self-check messages, and family-ready taken notices for KakaoTalk.

Use this agent when a small cafe owner wants to collect customer questions and receive one polite short reply draft in Notion.

Use this agent when apartment staff or caregivers need to check medication schedules from Google Sheets and produce a family-friendly KakaoTalk message plus a short PPT draft.

Use this agent when an internal lab team needs to turn local lab notes and .agentlas manifests into a public-safe marketplace listing, release bundle, scorecard, evidence pack, setup guide, and local CLI regression suite.

Use this agent when an evaluation engineer needs a privacy-preserving research dossier from GitHub issues and human feedback files; it produces executive-summary-first Markdown dossiers and runnable deterministic test fixtures.

Use this agent when you need to clean run logs and feedback files for private review; it produces sanitized local reports, regression evidence, and public-safe packaging checks.

Use this agent when you need to clean, organize, anonymize, and evaluate multimodal safety research material; it produces a short privacy-safe decision list, source register, failure summary, and next-run checklist.

Use this agent when a research lab needs to turn prototype notes, prompt corpora, threat models, and red-team transcripts into a public-safe marketplace listing with evidence, setup guides, and a reproducible local regression suite.

Use this agent when product requirements must become a marketplace-ready agent bundle with tool adapters, runtime tests, deployment checklists, safety boundaries, and rollback notes.

Use this agent when you have product requirements and existing agent evidence; it produces tool adapter plans, CI-ready runtime tests, deployment checklists, scorecards, and public-safe marketplace packaging.

Use this agent when you need CI-ready adversarial evaluation suites for tool-using LLM agents; it produces reproducible Markdown scorecards saved in the repo.

Use this agent when reviewing image-and-text safety research sources; it produces a short decision memo, a freshness-checked findings ledger, quarantine actions, and a day-one stale-finding smoke test.

Use this agent when an internal red-team lead needs to turn Agentlas manifests, exported ZIPs, or manual answers into a reproducible adversarial eval suite and executive scorecard for research or analysis agents.

Use this agent when you need a production-grade review of generated agent template or prompt-heavy repos; it produces a strict Markdown audit dossier, checklist summary, team message, and runnable deterministic test fixtures.

Use this agent when you have a lab prototype and source evidence; it produces a public-safe GitHub-ready marketplace package, setup guides, smoke tests, adversarial evaluations, telemetry notes, rollback plan, Markdown report, and JSON scorecard.

Use this agent when you need to turn notes, spreadsheets, docs, prior evaluation results, and failing examples into privacy-preserving feedback and model evaluation reports with audit trails and machine-readable manifests.

Use this agent when you need to analyze image-text safety findings from GitHub issues and product specs; it produces a Markdown research dossier, an updated memory ledger, runnable fixtures, adversarial evals, and a JSON readiness scorecard.

Use this agent when you have a vague agent idea and need a traceable research-grade report, build plan, eval harness outline, safety package, and machine-readable manifests.

Use this agent when you need to compare GitHub issues, product specs, and run telemetry against a research memory ledger; it produces a local Markdown report, JSON scorecard, trace evidence, and stop decision for stale or unsafe findings.

Use this agent when you need to review a generated agent repo before publishing; it produces a short pass/fail checklist, detailed findings, JSON results, and optional Slack delivery.

Use this agent when you need to turn one product requirements file into tool adapters, runtime tests, deployment checklists, and CI-ready evaluation reports.

Use this agent when reviewing new adversarial benchmark runs; it produces a traceable freshness report, conflict ledger, and machine-readable evidence manifest.

Use this agent when a lab team needs to compare Claude and Codex run methods on the same research tasks; it produces traceable JSON scorecards, CSV evidence tables, and a Markdown report.

Use this agent when you need to turn a lab prototype into a public-safe marketplace-ready agent bundle with evidence, setup guides, smoke tests, rollback notes, and safety boundaries.

Use this agent when a solo researcher has a vague agent idea or an existing agent repo; it produces a research-grade spec, implementation plan, eval harness, runnable repo draft, and JSON scorecard.

Use this agent when you need to stress-test autonomous agents across tool calls, memory, rollback, and budget exhaustion; it produces an executive Markdown dossier and runnable deterministic test fixtures.

Use this agent when you need to clean prompt corpora and feedback artifacts into privacy-preserving local evaluation evidence; it produces reports, scorecards, traces, adversarial results, and public-safe release notes.

Use this agent when you need to compare Claude, Codex, Gemini, and API runtimes; it produces reproducible local regression evidence with quality scores, cost traces, safety notes, and marketplace readiness results.

Use this agent when you need to review one generated agent repo with logs and model traces; it produces an evidence-first Markdown dossier, runnable fixtures, and an optional Slack summary.

Use this agent when you need to stress-test autonomous agents against tool failures, memory mistakes, rollback behavior, and budget exhaustion; it produces a pass/fail JSON scorecard, a Markdown report, and an optional Slack summary.

Use this agent when you need a read-only production review of one generated agent repo; it produces a JSON scorecard, Markdown report, and optional Slack summary.

Use this agent when research teams need a reproducible, privacy-safe plan and runnable fixtures for human feedback and model evaluation pipelines from local specs, telemetry exports, and approved GitHub or Drive source material.

Use this agent when you maintain open-source agent repositories and need adversarial eval suites plus reproducible marketplace-readiness scorecards from local generated repos, logs, and traces.

Use this agent when an AI engineering lab prototype must be turned into a public-safe marketplace listing; it produces a traceable readiness report, machine-readable manifests, setup guides, smoke tests, adversarial evals, and delivery evidence.

Use this agent when you need to turn product requirements, lab notes, Agentlas manifests, and exported ZIPs into deterministic local regression tests, deployment checklists, and public-safe marketplace packaging reports.

Use this agent when you need to review a generated agent repo; it produces a pass/fail Markdown dossier and runnable fixtures that expose hidden assumptions, missing setup, security risks, and weak tests.

Use this agent when you need to stress-test a local autonomous agent against tool use, memory behavior, rollback expectations, and budget exhaustion; it produces a traceable report, machine-readable manifests, and a public-safe marketplace package.

Use this agent when product requirements must be converted into tool adapter files, deterministic runtime tests, deployment checklists, and public-safe release evidence.

Use this agent when a solo builder needs to review a generated agent repo; it produces an executive-first Markdown dossier, deterministic smoke-test fixtures, adversarial eval notes, and optional Slack alerts.

Use this agent when you need to stress-test autonomous agents against tool call failures, memory drift, rollback mistakes, and spending limits; it produces executive markdown reports, JSON results, and optional Google Sheets summaries.

Use this agent when a vague agent idea must become a reproducible repo with a research-grade spec, implementation plan, eval harness, smoke tests, traceable report, and marketplace-ready manifests.

Use this agent when you need prompt-injection and secret-exfiltration eval suites for LLM apps; it produces reproducible JSON scorecards, Markdown reports, trace evidence, and marketplace readiness checks.

Use this agent when you need to review one generated agent repo for hidden assumptions, missing setup, security risks, weak tests, and marketplace readiness; it produces a short-verdict-first Markdown report and optional Slack summary.

Use this agent when you need to audit one prompt compiler project for schema drift, jailbreak surface, silent fallback behavior, and release readiness; it produces a traceable report, JSON scorecard, risk table, and marketplace-safe manifest.

Use this agent when reviewing new alignment research from generated agent repos and run logs; it produces short decision memos, JSON evaluation results, stale-finding alerts, and an optional Slack summary.

Use this agent when a solo developer needs to compare Claude, Codex, Gemini, and API run methods on one task and receive a traceable scorecard with quality, cost, safety, and setup evidence.

Use this agent when you need a CI-ready evaluation package for prompt-injection, secret exfiltration, human feedback, and model evaluation; it produces sanitized reports, deterministic tests, risk decisions, and rollback guidance.

Use this agent when you need a local CLI regression suite that checks human feedback and model evaluation data for quality, privacy, reproducibility, and public-safe release readiness; it produces Markdown reports and CSV findings.

Use this agent when you need a read-only review of generated agent repositories; it produces an executive-first security report, traceable findings, and machine-readable review manifests.

Use this agent when you need to review new AI infrastructure research without letting stale findings contaminate the current run; it produces a Markdown research dossier, updated ledger entries, delivery message, and runnable test fixtures.

Use this agent when you need to stress-test an autonomous agent's tool-use failures, memory behavior, rollback path, and budget exhaustion; it produces a research-style Markdown dossier, runnable test fixtures, Google Sheets summary, and Slack delivery note.

Use this agent when an internal safety team needs to clean, de-identify, package, and score human feedback, run logs, and model traces into GitHub-ready evaluation datasets with a JSON scorecard.

Use this agent when agent builders need a production-readiness review of a generated agent repo; it produces a Markdown release checklist and optional Slack summary covering assumptions, setup gaps, safety risks, tests, telemetry, and rollback readiness.

Use this agent when you need to turn one open-source agent project into a public-safe marketplace package; it produces a Markdown research dossier, setup guides, evidence index, and runnable test fixtures.

Use this agent when you need to audit prompt compiler changes; it produces a pass/fail JSON scorecard and Markdown report covering schema drift, jailbreak surface, and silent fallback behavior.

Use this agent when an internal safety team needs to audit prompt compiler code, specs, issues, and telemetry; it produces a repo JSON scorecard, a Markdown audit report, and an optional GitHub issue comment.

Use this agent when you need to compare Claude, Codex, Gemini, and API run results; it produces a traceable JSON scorecard and Markdown evidence report.

Use this agent when you need to audit prompt compiler inputs and outputs for schema drift, jailbreak exposure, and silent fallback behavior; it produces a short pass/fail checklist, JSON scorecard, traces, and readiness verdict.

Use this agent when evaluating prompt corpora and threat models; it produces fresh decision memos, a research memory ledger, stale-finding warnings, and CI-ready evaluation evidence.

Use this agent when a solo engineer needs traceable local comparisons of Claude and Codex output quality, cost, failures, and marketplace readiness; it produces a short pass/fail report, JSON scorecard, evidence bundle, and known-failures log.

Use this agent when a red-team lead needs to audit one prompt compiler; it produces a Markdown research dossier, runnable fixtures, drift evidence, adversarial checks, and a clear stop-or-rollback record.

Use this agent when you have product requirements and source notes; it produces adapter plans, tool contracts, runtime tests, deployment checklists, adversarial evals, and a JSON scorecard as draft repo files.

Use this agent when you need adversarial evaluation suites for document QA agents; it produces reproducible eval files, scorecards, and public-safe release packages.

Use this agent when you need to review one generated agent repo for hidden assumptions, missing setup, security risks, weak tests, and marketplace readiness; it produces pass/fail Markdown and JSON regression reports.

Use this agent when you maintain portable agent skills or adapters and need a local CLI suite that stress-tests tool use, memory behavior, rollback, spending limits, and marketplace safety, producing JSON regression reports plus optional Slack summaries.

Use this agent when you need to design adversarial evaluation suites for document or web agents and produce reproducible local scorecards with trace and cost evidence.

Use this agent when an LLM observability founder needs to refresh research memory from GitHub issues, product specs, and trace or cost evidence; it produces a stale-aware Markdown research dossier, JSON scorecard, and runnable test fixtures.

Use this agent when an LLM observability founder needs to turn local Agentlas manifests and exported ZIPs into a public-safe marketplace bundle with evidence, setup guides, tests, and a readiness score.

Use this agent when a product manager needs to turn a requirement plus prior test evidence into adapter plans, runtime tests, deployment checklists, and a traceable report bundle.

Use this agent when you need a CI-ready audit of prompt compiler source files; it produces Markdown and JSON reports for schema drift, jailbreak surface, and review-stopping failures.

Use this agent when an internal research team needs to design adversarial evaluation suites for search or research agents and produce reproducible JSON, Markdown, and CSV scorecards.

Use this agent when you have a vague agent idea and existing source files; it produces a research-grade spec, implementation plan, eval harness, runnable repo plan, and public-safe marketplace bundle.

Use this agent when an internal evals team needs to audit prompt compiler outputs for schema drift, jailbreak surface, and silent fallback behavior; it produces an executive Markdown report, dashboard summary, risk score, and follow-up tasks.

Use this agent when a product requirement must become connector designs, tool contracts, runtime tests, deployment checklists, and a traceable report with machine-readable manifests.

Use this agent when you need to compare new research files against an append-only memory ledger; it produces a Markdown dossier, stale-claim findings, and runnable test fixtures.

Use this agent when an internal safety team needs to turn lab prototype materials into a public-safe marketplace listing with evidence, setup guides, checks, and stop conditions.

Use this agent when you need a local CLI review of generated agent repositories; it produces a human report, structured findings JSON, and a fix checklist for hidden assumptions, setup gaps, unsafe packaging, security risks, and weak tests.

Use this agent when you need to review paper notes, API docs, prior eval results, and failing examples; it produces a short Markdown research dossier that flags stale findings, conflicting evidence, and unsafe public-packaging risks.

Use this agent when you need reproducible prompt-injection and secret-exfiltration test suites for LLM apps; it produces deterministic scorecards, assumptions, evidence links, and safe remediation notes.

Use this agent when you need to turn an API tool lab prototype into a public-safe marketplace package; it produces a checklist-first traceable report, marketplace listing draft, setup guide, manifests, smoke-test evidence, adversarial eval summary, telemetry n

Use this agent when you need a local command-line suite that reads Agentlas manifests and exported ZIPs, designs adversarial evaluation tests, and produces reproducible short scorecards.

Use this agent when you need to turn prompt corpora into adversarial regression tests for RAG or search applications; it produces executive-first Markdown scorecards, structured JSON results, and an optional Slack summary.

Use this agent when you need to audit prompt compiler packages for schema drift, jailbreak exposure, and silent fallback behavior; it produces a traceable report, findings list, and machine-readable audit manifest.

Use this agent when reviewing risky prompt examples and threat models; it produces a privacy-preserving executive research dossier plus runnable deterministic test fixtures.

Use this agent when an LLM observability founder needs to stress-test autonomous agents; it produces a reproducible local CLI smoke suite, adversarial eval results, trace and cost evidence, rollback checks, and marketplace readiness scorecards.

Use this agent when a solo developer needs to turn product requirements into a marketplace-ready adapter bundle with run-method tests, deployment checklists, smoke tests, adversarial evals, telemetry evidence, rollback behavior, and public-safe packaging.

Use this agent when a research team needs to turn paper notes, API docs, prior eval results, and failing examples into privacy-preserving feedback datasets, evaluation fixtures, and a research-style Markdown dossier.

Use this agent when open-source maintainers need to compare Claude, Codex, Gemini, and API-based agents on the same prompt and threat-model materials; it produces a traceable JSON scorecard, Markdown report, and optional Slack summary.

Use this agent when AI infrastructure engineers need to turn prototype notes, API docs, prototype files, eval results, and failing examples into a public-safe marketplace-ready agent bundle with evidence, setup guides, tests, safety boundaries, and rollback be

Use this agent when you need to stress-test autonomous agents with source corpora and transcripts; it produces traceable Markdown reports, JSON manifests, Slack summaries, and pass/fail evidence.

Use this agent when mechanistic interpretability research needs a fresh-run staleness check; it produces a traceable report, active-findings ledger, and machine-readable manifest.

Use this agent when auditing generated agent repos, run logs, and model traces; it produces a GitHub-ready security audit package plus a JSON scorecard.

Use this agent when you have a vague agent idea and source material; it produces a research-grade spec, implementation plan, eval harness, runnable repo plan, safety model, and marketplace-ready bundle.

Use this agent when a research lab needs to turn prompt corpora, threat models, and red-team transcripts into a public-safe GitHub-ready marketplace repo with evidence, setup guides, and a JSON scorecard.

Use this agent when you need to compare Claude Code, Codex, Gemini, and API-based run methods; it produces a traceable recommendation, score table, and machine-readable manifest for choosing the best runtime for a task.

Use this agent when you need adversarial evaluation suites for code agents; it reads local adapter contracts and setup guides, runs deterministic checks, and produces reproducible executive scorecards.

Use this agent when you need to compare Claude, Codex, Gemini, and API model runs; it produces traceable quality, cost, disagreement, and audit evidence reports.

Use this agent when you need to turn paper notes, API docs, prior eval results, and failing image-text examples into clean privacy-preserving evaluation datasets with a research report.

Use this agent when you need to ingest local AI infrastructure research notes, isolate projects, refresh stale findings with newer evidence, and produce a CI-ready pass/fail evaluation package.

Use this agent when you have one vague agent idea and existing generated repos, run logs, or model traces; it produces a research-grade Markdown dossier, implementation plan, eval harness, runnable repo scaffold, JSON scorecard, trace evidence, and public-safe

Use this agent when you need a read-only production readiness review of a generated agent repo; it produces a Markdown research dossier, JSON scorecard, runnable smoke fixtures, adversarial eval cases, and a publish-or-block decision.

Use this agent when you need a local regression suite that audits prompt compilers for schema drift, jailbreak surface, silent fallback behavior, and marketplace readiness from local contracts, guides, Git history, and saved CI logs.

Use this agent when you need a read-only safety review of coding agents; it produces an executive report on tool use, memory behavior, rollback readiness, spending limits, and refusal boundaries.

정리되지 않은 사용자 버그 제보를 재현 단계, 기대/실제 동작, 심각도, 추가 정보 요청이 포함된 Notion 개발 티켓으로 자동 생성합니다.

"Plan my brand's Instagram content every Monday and write 4 carousel post drafts plus captions for the week."

4 carousel drafts, captions, and a weekly posting plan

"Read incoming customer support tickets, classify the issue, draft a calm reply, and add an internal note about the underlying bug."

classified tickets, reply drafts, and internal bug notes

"Research a topic across web sources, academic papers, industry reports, and internal documents, then produce a consulting-grade long-form proposal with an evidence matrix, recommendation, roadmap, risks, and layout-ready Word/PPT output notes."

source log, evidence matrix, long-form proposal, and layout-ready report/deck outline

"Review the diff and produce a punch list — must-fix, should-fix, nice-to-have. Skip bikeshedding."

must-fix, should-fix, and nice-to-have review list

"Write 3 hook variants and a 45-second script for a YouTube Short on a topic I give you."

3 hooks and a 45-second script

"Every morning, summarize last 24h activity from GitHub and Slack, flag anything stalled, and post a digest to a Slack channel."

24-hour digest, blockers, and Slack-ready summary

"Summarize an academic paper for a non-specialist: claim, evidence quality, why it matters, and the one chart worth looking at."

claim, evidence quality, business meaning, and chart callout

"Generate launch comms for a new feature: a 280-char tweet, a 3-paragraph blog intro, an email subject, and a Product Hunt tagline."

tweet, blog intro, email subject, and Product Hunt tagline

"Find 20 qualified B2B leads for my product, write a one-line reason each account fits, add the source link, and draft a short first outreach email."

qualified lead table, first-line personalization, and source links

"Turn a client meeting transcript into decisions, open questions, owner-by-owner next steps, and a polished follow-up email."

decisions, owners, next steps, and follow-up email

"Read recent customer reviews or feedback notes, group them into themes, pull representative quotes, score severity, and suggest product fixes."

themes, quotes, severity, and roadmap suggestions

"Draft a monthly investor update from my notes: wins, metrics, product progress, risks, asks, and what we need help with."

monthly investor email with wins, metrics, risks, and asks

"Review an unpaid invoice list every Friday, group accounts by days overdue, draft polite follow-up emails, and flag which ones need escalation."

overdue summary, polite reminder drafts, and escalation list

"Prepare an App Store release brief for a new mobile build: release notes, keyword angle, screenshot caption ideas, and review-risk checklist."

release notes, keyword angle, screenshot copy, and risk notes

"Convert a messy user bug report into a clean engineering ticket with repro steps, expected behavior, actual behavior, severity, and missing info."

repro steps, severity, likely owner, and ticket body

"Answer a vendor security questionnaire from our policy notes, mark unknown answers clearly, cite the evidence file, and list follow-up owners."

questionnaire answers, evidence links, and unknowns

"Turn my class notes and upcoming exam topics into a weekly study plan, weak-topic checklist, practice questions, and reminders written for a student."

study plan, weak-topic list, practice prompts, and check-in reminders

"Read customer messages for my local shop, group them by intent, draft friendly replies, and suggest one simple follow-up or promo message."

polite replies, promo notes, and follow-up messages

"Turn clinic visit notes into patient-friendly follow-up instructions, appointment reminders, and a list of missing information to confirm."

appointment follow-up, patient-friendly instructions, and missing-info list

"Convert citizen or community requests into a clear summary, category, urgency level, missing information, and a plain-language response draft."

request summary, category, urgency, and response draft

"Turn a legal intake note into a neutral case summary, timeline, evidence checklist, and follow-up questions without giving legal advice."

case summary, timeline, evidence list, and follow-up questions

"Convert interview notes into a candidate summary, strengths, risk flags, scorecard bullets, and a next-step email draft."

candidate summary, strengths, risks, and next-step email

"Summarize my monthly income and expenses, flag unusual subscriptions, draft payment follow-ups, and create a simple next-month checklist."

monthly cash summary, subscription review, and follow-up checklist

"Build a travel itinerary from my notes, compare vendors, list booking questions, and create a practical packing and day-by-day plan."

itinerary, vendor comparison, booking questions, and packing notes

"Turn property notes into listing copy, likely buyer questions, viewing follow-up messages, and missing details to confirm."

listing copy, buyer questions, and follow-up messages

"Adapt my product launch notes for a new country: local examples, language tone risks, currency/trust notes, and revised launch copy."

localized examples, currency/trust notes, and launch copy risks

"Review a screen's copy and controls for accessibility, screen-reader labels, mobile density, unclear actions, and safer button wording."

screen-reader labels, unclear actions, and mobile-density fixes

"Turn a SaaS pricing page into a plain explanation: what each plan buys, rough usage examples, currency/tax caveats, and checkout trust questions."

price explanation, usage estimate, currency note, and trust checklist